Merge tag 'drm-intel-next-2020-03-13' of git://anongit.freedesktop.org/drm/drm-intel into drm-next

UAPI Changes:

On i915 we have a new UAPI to allow userspace to specify CS ring buffer size on
construction (I915_CONTEXT_PARAM_RINGSIZE) and also new sysfs entries exposing
various engine properties

GVT Changes:

VFIO edid getting expanded to all platforms and a big cleanup around attr
group, unused vblank complete, kvmgt, Intel engine and dev_priv usages.

i915 Changes:

- new UAPI to allow userspace to specify CS ring buffer size on construction
  (I915_CONTEXT_PARAM_RINGSIZE) -  (Chris)
- New sysfs entries exposing various engine properties (Chris)
- Tiger Lake is out of require_force_probe protection (Jose)
- Changes in many places around active requests, reset and heartbeat (Chris)
- Stop assigning drm-dev_private pointer (Jani)
- Many code refactor in many places, including intel_modeset_init,
  increasing use of intel_uncore_*, vgpu, and gvt stuff (Jani)
- Fixes around display pipe iterators (Anshuman)
- Tigerlake enabling work (Matt Ropper, Matt Atwood, Ville, Lucas, Daniele,
  Jose, Anusha, Vivek, Swathi, Caz. Kai)
- Code clean-up like reducing use of drm/i915_drv.h, removing unused
  registers, removing garbage warns, and some other code polishing (Jani, Lucas,
  Ville)
- Selftests fixes, improvements and additions (Chris, Dan, Aditya, Matt Auld)
- Fix plane possible_crtcs bit mask (Anshuman)
- Fixes and cleanup on GLK pre production identification and w/a (Ville)
- Fix display orientation on few cases (Hans, Ville)
- dbuf clean-up and improvements for slice arrays handling (Ville)
- Improvement around min cdclk calculation (Stanislav)
- Fixes and refactor around display PLLs (Imre)
- Other execlists and perf fixes (Chris)
- Documentation fixes (Jani, Chris)
- Fix build issue (Anshuman)
- Many more fixes around the locking mechanisms (Chris)
- Other fixes and debugability info around preemption (Chris, Tvrtko)
- Add mechanism to submit a context WA on ring submission (Mika)
- Clear all Eu/L3 resitual context (Prathap)
- More changes around local memory (Abdiel, Matt, Chris)
- Fix RPS (Chris)
- DP MST fix (Lyude)
- Display FBC fixes (Jose, RK)
- debugfs cleanup (Tvrtko)
- More convertion towards drm_debive based loggin (Wambui, Ram)
- Avoid potential buffer overflow (Takashi)
- Ice Lake and Elkhart Lake workarounds (Matt Roper)

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

From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200314001535.GA2969344@intel.com

166 files changed, 6868 insertions, 4123 deletions

diff --git a/drivers/gpu/drm/i915/Kconfig.profile b/drivers/gpu/drm/i915/Kconfig.profile
index c280b6ae38eb..0bfd276c19fe 100644
--- a/drivers/gpu/drm/i915/Kconfig.profile
+++ b/drivers/gpu/drm/i915/Kconfig.profile
@@ -20,6 +20,9 @@ config DRM_I915_HEARTBEAT_INTERVAL
 	  check the health of the GPU and undertake regular house-keeping of
 	  internal driver state.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/heartbeat_interval_ms
+
 	  May be 0 to disable heartbeats and therefore disable automatic GPU
 	  hang detection.
 
@@ -33,11 +36,18 @@ config DRM_I915_PREEMPT_TIMEOUT
 	  expires, the HW will be reset to allow the more important context
 	  to execute.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/preempt_timeout_ms
+
 	  May be 0 to disable the timeout.
 
-config DRM_I915_SPIN_REQUEST
-	int "Busywait for request completion (us)"
-	default 5 # microseconds
+	  The compiled in default may get overridden at driver probe time on
+	  certain platforms and certain engines which will be reflected in the
+	  sysfs control.
+
+config DRM_I915_MAX_REQUEST_BUSYWAIT
+	int "Busywait for request completion limit (ns)"
+	default 8000 # nanoseconds
 	help
 	  Before sleeping waiting for a request (GPU operation) to complete,
 	  we may spend some time polling for its completion. As the IRQ may
@@ -45,6 +55,9 @@ config DRM_I915_SPIN_REQUEST
 	  check if the request will complete in the time it would have taken
 	  us to enable the interrupt.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/max_busywait_duration_ns
+
 	  May be 0 to disable the initial spin. In practice, we estimate
 	  the cost of enabling the interrupt (if currently disabled) to be
 	  a few microseconds.
@@ -60,6 +73,9 @@ config DRM_I915_STOP_TIMEOUT
 	  that the reset itself may take longer and so be more disruptive to
 	  interactive or low latency workloads.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/stop_timeout_ms
+
 config DRM_I915_TIMESLICE_DURATION
 	int "Scheduling quantum for userspace batches (ms, jiffy granularity)"
 	default 1 # milliseconds
@@ -73,4 +89,7 @@ config DRM_I915_TIMESLICE_DURATION
 	  is scheduled for execution for the timeslice duration, before
 	  switching to the next context.
 
+	  This is adjustable via
+	  /sys/class/drm/card?/engine/*/timeslice_duration_ms
+
 	  May be 0 to disable timeslicing.
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index bc28c31c4f78..9f887a86e555 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -47,6 +47,7 @@ i915-y += i915_drv.o \
 	  i915_sysfs.o \
 	  i915_utils.o \
 	  intel_device_info.o \
+	  intel_dram.o \
 	  intel_memory_region.o \
 	  intel_pch.o \
 	  intel_pm.o \
@@ -79,9 +80,11 @@ gt-y += \
 	gt/debugfs_gt.o \
 	gt/debugfs_gt_pm.o \
 	gt/gen6_ppgtt.o \
+	gt/gen7_renderclear.o \
 	gt/gen8_ppgtt.o \
 	gt/intel_breadcrumbs.o \
 	gt/intel_context.o \
+	gt/intel_context_param.o \
 	gt/intel_context_sseu.o \
 	gt/intel_engine_cs.o \
 	gt/intel_engine_heartbeat.o \
@@ -107,7 +110,8 @@ gt-y += \
 	gt/intel_rps.o \
 	gt/intel_sseu.o \
 	gt/intel_timeline.o \
-	gt/intel_workarounds.o
+	gt/intel_workarounds.o \
+	gt/sysfs_engines.o
 # autogenerated null render state
 gt-y += \
 	gt/gen6_renderstate.o \
diff --git a/drivers/gpu/drm/i915/display/icl_dsi.c b/drivers/gpu/drm/i915/display/icl_dsi.c
index d842e280699d..17cee6f80d8b 100644
--- a/drivers/gpu/drm/i915/display/icl_dsi.c
+++ b/drivers/gpu/drm/i915/display/icl_dsi.c
@@ -599,13 +599,13 @@ static void gen11_dsi_gate_clocks(struct intel_encoder *encoder)
 	u32 tmp;
 	enum phy phy;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 	tmp = intel_de_read(dev_priv, ICL_DPCLKA_CFGCR0);
 	for_each_dsi_phy(phy, intel_dsi->phys)
 		tmp |= ICL_DPCLKA_CFGCR0_DDI_CLK_OFF(phy);
 
 	intel_de_write(dev_priv, ICL_DPCLKA_CFGCR0, tmp);
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static void gen11_dsi_ungate_clocks(struct intel_encoder *encoder)
@@ -615,13 +615,13 @@ static void gen11_dsi_ungate_clocks(struct intel_encoder *encoder)
 	u32 tmp;
 	enum phy phy;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 	tmp = intel_de_read(dev_priv, ICL_DPCLKA_CFGCR0);
 	for_each_dsi_phy(phy, intel_dsi->phys)
 		tmp &= ~ICL_DPCLKA_CFGCR0_DDI_CLK_OFF(phy);
 
 	intel_de_write(dev_priv, ICL_DPCLKA_CFGCR0, tmp);
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static void gen11_dsi_map_pll(struct intel_encoder *encoder,
@@ -633,7 +633,7 @@ static void gen11_dsi_map_pll(struct intel_encoder *encoder,
 	enum phy phy;
 	u32 val;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 
 	val = intel_de_read(dev_priv, ICL_DPCLKA_CFGCR0);
 	for_each_dsi_phy(phy, intel_dsi->phys) {
@@ -652,7 +652,7 @@ static void gen11_dsi_map_pll(struct intel_encoder *encoder,
 
 	intel_de_posting_read(dev_priv, ICL_DPCLKA_CFGCR0);
 
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static void
@@ -1350,15 +1350,15 @@ static void gen11_dsi_get_timings(struct intel_encoder *encoder,
 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 drm_i915_private *i915 = to_i915(encoder->base.dev);
 	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(encoder);
 
 	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->port_clock = intel_dpll_get_freq(i915,
+						      pipe_config->shared_dpll);
 
 	pipe_config->hw.adjusted_mode.crtc_clock = intel_dsi->pclk;
 	if (intel_dsi->dual_link)
diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.c b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
index c86d7a35c816..457b258683d3 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic_plane.c
+++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
@@ -133,15 +133,37 @@ intel_plane_destroy_state(struct drm_plane *plane,
 	kfree(plane_state);
 }
 
+unsigned int intel_plane_pixel_rate(const struct intel_crtc_state *crtc_state,
+				    const struct intel_plane_state *plane_state)
+{
+	unsigned int src_w, src_h, dst_w, dst_h;
+	unsigned int pixel_rate = crtc_state->pixel_rate;
+
+	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);
+	dst_h = min(src_h, dst_h);
+
+	return DIV_ROUND_UP_ULL(mul_u32_u32(pixel_rate, src_w * src_h),
+				dst_w * dst_h);
+}
+
 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->hw.fb;
 	unsigned int cpp;
+	unsigned int pixel_rate;
 
 	if (!plane_state->uapi.visible)
 		return 0;
 
+	pixel_rate = intel_plane_pixel_rate(crtc_state, plane_state);
+
 	cpp = fb->format->cpp[0];
 
 	/*
@@ -153,7 +175,7 @@ unsigned int intel_plane_data_rate(const struct intel_crtc_state *crtc_state,
 	if (fb->format->is_yuv && fb->format->num_planes > 1)
 		cpp *= 4;
 
-	return cpp * crtc_state->pixel_rate;
+	return pixel_rate * cpp;
 }
 
 int intel_plane_calc_min_cdclk(struct intel_atomic_state *state,
diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.h b/drivers/gpu/drm/i915/display/intel_atomic_plane.h
index 2bcf15e34728..a6bbf42bae1f 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic_plane.h
+++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.h
@@ -18,6 +18,9 @@ struct intel_plane_state;
 
 extern const struct drm_plane_helper_funcs intel_plane_helper_funcs;
 
+unsigned int intel_plane_pixel_rate(const struct intel_crtc_state *crtc_state,
+				    const struct intel_plane_state *plane_state);
+
 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,
diff --git a/drivers/gpu/drm/i915/display/intel_audio.c b/drivers/gpu/drm/i915/display/intel_audio.c
index 30fb7c887ff0..62f234f641de 100644
--- a/drivers/gpu/drm/i915/display/intel_audio.c
+++ b/drivers/gpu/drm/i915/display/intel_audio.c
@@ -149,6 +149,10 @@ static const struct {
 	{ 74250, AUD_CONFIG_PIXEL_CLOCK_HDMI_74250 },
 	{ 148352, AUD_CONFIG_PIXEL_CLOCK_HDMI_148352 },
 	{ 148500, AUD_CONFIG_PIXEL_CLOCK_HDMI_148500 },
+	{ 296703, AUD_CONFIG_PIXEL_CLOCK_HDMI_296703 },
+	{ 297000, AUD_CONFIG_PIXEL_CLOCK_HDMI_297000 },
+	{ 593407, AUD_CONFIG_PIXEL_CLOCK_HDMI_593407 },
+	{ 594000, AUD_CONFIG_PIXEL_CLOCK_HDMI_594000 },
 };
 
 /* HDMI N/CTS table */
@@ -234,6 +238,7 @@ static const struct hdmi_aud_ncts hdmi_aud_ncts_36bpp[] = {
 /* get AUD_CONFIG_PIXEL_CLOCK_HDMI_* value for mode */
 static u32 audio_config_hdmi_pixel_clock(const struct intel_crtc_state *crtc_state)
 {
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	const struct drm_display_mode *adjusted_mode =
 		&crtc_state->hw.adjusted_mode;
 	int i;
@@ -243,6 +248,9 @@ static u32 audio_config_hdmi_pixel_clock(const struct intel_crtc_state *crtc_sta
 			break;
 	}
 
+	if (INTEL_GEN(dev_priv) < 12 && adjusted_mode->crtc_clock > 148500)
+		i = ARRAY_SIZE(hdmi_audio_clock);
+
 	if (i == ARRAY_SIZE(hdmi_audio_clock)) {
 		DRM_DEBUG_KMS("HDMI audio pixel clock setting for %d not found, falling back to defaults\n",
 			      adjusted_mode->crtc_clock);
@@ -844,7 +852,7 @@ static void glk_force_audio_cdclk(struct drm_i915_private *dev_priv,
 	struct intel_crtc *crtc;
 	int ret;
 
-	crtc = intel_get_crtc_for_pipe(dev_priv, PIPE_A);
+	crtc = intel_get_first_crtc(dev_priv);
 	if (!crtc)
 		return;
 
diff --git a/drivers/gpu/drm/i915/display/intel_bios.c b/drivers/gpu/drm/i915/display/intel_bios.c
index 2049cf5b54f3..839124647202 100644
--- a/drivers/gpu/drm/i915/display/intel_bios.c
+++ b/drivers/gpu/drm/i915/display/intel_bios.c
@@ -26,7 +26,6 @@
  */
 
 #include <drm/drm_dp_helper.h>
-#include <drm/i915_drm.h>
 
 #include "display/intel_display.h"
 #include "display/intel_display_types.h"
diff --git a/drivers/gpu/drm/i915/display/intel_bios.h b/drivers/gpu/drm/i915/display/intel_bios.h
index c17199caeff8..e29e79faa01b 100644
--- a/drivers/gpu/drm/i915/display/intel_bios.h
+++ b/drivers/gpu/drm/i915/display/intel_bios.h
@@ -32,8 +32,6 @@
 
 #include <linux/types.h>
 
-#include <drm/i915_drm.h>
-
 struct drm_i915_private;
 struct intel_crtc_state;
 struct intel_encoder;
diff --git a/drivers/gpu/drm/i915/display/intel_cdclk.c b/drivers/gpu/drm/i915/display/intel_cdclk.c
index 0741d643455b..979a0241fdcb 100644
--- a/drivers/gpu/drm/i915/display/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/display/intel_cdclk.c
@@ -1868,6 +1868,8 @@ static void intel_set_cdclk(struct drm_i915_private *dev_priv,
 			    const struct intel_cdclk_config *cdclk_config,
 			    enum pipe pipe)
 {
+	struct intel_encoder *encoder;
+
 	if (!intel_cdclk_changed(&dev_priv->cdclk.hw, cdclk_config))
 		return;
 
@@ -1876,8 +1878,28 @@ static void intel_set_cdclk(struct drm_i915_private *dev_priv,
 
 	intel_dump_cdclk_config(cdclk_config, "Changing CDCLK to");
 
+	/*
+	 * Lock aux/gmbus while we change cdclk in case those
+	 * functions use cdclk. Not all platforms/ports do,
+	 * but we'll lock them all for simplicity.
+	 */
+	mutex_lock(&dev_priv->gmbus_mutex);
+	for_each_intel_dp(&dev_priv->drm, encoder) {
+		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
+
+		mutex_lock_nest_lock(&intel_dp->aux.hw_mutex,
+				     &dev_priv->gmbus_mutex);
+	}
+
 	dev_priv->display.set_cdclk(dev_priv, cdclk_config, pipe);
 
+	for_each_intel_dp(&dev_priv->drm, encoder) {
+		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
+
+		mutex_unlock(&intel_dp->aux.hw_mutex);
+	}
+	mutex_unlock(&dev_priv->gmbus_mutex);
+
 	if (drm_WARN(&dev_priv->drm,
 		     intel_cdclk_changed(&dev_priv->cdclk.hw, cdclk_config),
 		     "cdclk state doesn't match!\n")) {
diff --git a/drivers/gpu/drm/i915/display/intel_color.c b/drivers/gpu/drm/i915/display/intel_color.c
index 36dd52d2a9ee..c1cce93a1c25 100644
--- a/drivers/gpu/drm/i915/display/intel_color.c
+++ b/drivers/gpu/drm/i915/display/intel_color.c
@@ -348,48 +348,56 @@ static void icl_load_csc_matrix(const struct intel_crtc_state *crtc_state)
 		       crtc_state->csc_mode);
 }
 
-/*
- * Set up the pipe CSC unit on CherryView.
- */
-static void cherryview_load_csc_matrix(const struct intel_crtc_state *crtc_state)
+static void chv_load_cgm_csc(struct intel_crtc *crtc,
+			     const struct drm_property_blob *blob)
 {
-	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_color_ctm *ctm = blob->data;
 	enum pipe pipe = crtc->pipe;
+	u16 coeffs[9];
+	int i;
 
-	if (crtc_state->hw.ctm) {
-		const struct drm_color_ctm *ctm = crtc_state->hw.ctm->data;
-		u16 coeffs[9] = {};
-		int i;
-
-		for (i = 0; i < ARRAY_SIZE(coeffs); i++) {
-			u64 abs_coeff =
-				((1ULL << 63) - 1) & ctm->matrix[i];
-
-			/* Round coefficient. */
-			abs_coeff += 1 << (32 - 13);
-			/* Clamp to hardware limits. */
-			abs_coeff = clamp_val(abs_coeff, 0, CTM_COEFF_8_0 - 1);
-
-			/* Write coefficients in S3.12 format. */
-			if (ctm->matrix[i] & (1ULL << 63))
-				coeffs[i] = 1 << 15;
-			coeffs[i] |= ((abs_coeff >> 32) & 7) << 12;
-			coeffs[i] |= (abs_coeff >> 20) & 0xfff;
-		}
+	for (i = 0; i < ARRAY_SIZE(coeffs); i++) {
+		u64 abs_coeff = ((1ULL << 63) - 1) & ctm->matrix[i];
+
+		/* Round coefficient. */
+		abs_coeff += 1 << (32 - 13);
+		/* Clamp to hardware limits. */
+		abs_coeff = clamp_val(abs_coeff, 0, CTM_COEFF_8_0 - 1);
 
-		intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF01(pipe),
-			       coeffs[1] << 16 | coeffs[0]);
-		intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF23(pipe),
-			       coeffs[3] << 16 | coeffs[2]);
-		intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF45(pipe),
-			       coeffs[5] << 16 | coeffs[4]);
-		intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF67(pipe),
-			       coeffs[7] << 16 | coeffs[6]);
-		intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF8(pipe), coeffs[8]);
+		coeffs[i] = 0;
+
+		/* Write coefficients in S3.12 format. */
+		if (ctm->matrix[i] & (1ULL << 63))
+			coeffs[i] |= 1 << 15;
+
+		coeffs[i] |= ((abs_coeff >> 32) & 7) << 12;
+		coeffs[i] |= (abs_coeff >> 20) & 0xfff;
 	}
 
-	intel_de_write(dev_priv, CGM_PIPE_MODE(pipe), crtc_state->cgm_mode);
+	intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF01(pipe),
+		       coeffs[1] << 16 | coeffs[0]);
+	intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF23(pipe),
+		       coeffs[3] << 16 | coeffs[2]);
+	intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF45(pipe),
+		       coeffs[5] << 16 | coeffs[4]);
+	intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF67(pipe),
+		       coeffs[7] << 16 | coeffs[6]);
+	intel_de_write(dev_priv, CGM_PIPE_CSC_COEFF8(pipe),
+		       coeffs[8]);
+}
+
+/* convert hw value with given bit_precision to lut property val */
+static u32 intel_color_lut_pack(u32 val, int bit_precision)
+{
+	u32 max = 0xffff >> (16 - bit_precision);
+
+	val = clamp_val(val, 0, max);
+
+	if (bit_precision < 16)
+		val <<= 16 - bit_precision;
+
+	return val;
 }
 
 static u32 i9xx_lut_8(const struct drm_color_lut *color)
@@ -399,6 +407,13 @@ static u32 i9xx_lut_8(const struct drm_color_lut *color)
 		drm_color_lut_extract(color->blue, 8);
 }
 
+static void i9xx_lut_8_pack(struct drm_color_lut *entry, u32 val)
+{
+	entry->red = intel_color_lut_pack(REG_FIELD_GET(LGC_PALETTE_RED_MASK, val), 8);
+	entry->green = intel_color_lut_pack(REG_FIELD_GET(LGC_PALETTE_GREEN_MASK, val), 8);
+	entry->blue = intel_color_lut_pack(REG_FIELD_GET(LGC_PALETTE_BLUE_MASK, val), 8);
+}
+
 /* i965+ "10.6" bit interpolated format "even DW" (low 8 bits) */
 static u32 i965_lut_10p6_ldw(const struct drm_color_lut *color)
 {
@@ -415,48 +430,34 @@ static u32 i965_lut_10p6_udw(const struct drm_color_lut *color)
 		(color->blue >> 8);
 }
 
-static u32 ilk_lut_10(const struct drm_color_lut *color)
+static void i965_lut_10p6_pack(struct drm_color_lut *entry, u32 ldw, u32 udw)
 {
-	return drm_color_lut_extract(color->red, 10) << 20 |
-		drm_color_lut_extract(color->green, 10) << 10 |
-		drm_color_lut_extract(color->blue, 10);
+	entry->red = REG_FIELD_GET(PALETTE_RED_MASK, udw) << 8 |
+		REG_FIELD_GET(PALETTE_RED_MASK, ldw);
+	entry->green = REG_FIELD_GET(PALETTE_GREEN_MASK, udw) << 8 |
+		REG_FIELD_GET(PALETTE_GREEN_MASK, ldw);
+	entry->blue = REG_FIELD_GET(PALETTE_BLUE_MASK, udw) << 8 |
+		REG_FIELD_GET(PALETTE_BLUE_MASK, ldw);
 }
 
-/* Loads the legacy palette/gamma unit for the CRTC. */
-static void i9xx_load_luts_internal(const struct intel_crtc_state *crtc_state,
-				    const struct drm_property_blob *blob)
+static u16 i965_lut_11p6_max_pack(u32 val)
 {
-	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;
-
-	if (HAS_GMCH(dev_priv)) {
-		if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI))
-			assert_dsi_pll_enabled(dev_priv);
-		else
-			assert_pll_enabled(dev_priv, pipe);
-	}
-
-	if (blob) {
-		const struct drm_color_lut *lut = blob->data;
-
-		for (i = 0; i < 256; i++) {
-			u32 word = i9xx_lut_8(&lut[i]);
+	/* PIPEGCMAX is 11.6, clamp to 10.6 */
+	return clamp_val(val, 0, 0xffff);
+}
 
-			if (HAS_GMCH(dev_priv))
-				intel_de_write(dev_priv, PALETTE(pipe, i),
-					       word);
-			else
-				intel_de_write(dev_priv, LGC_PALETTE(pipe, i),
-					       word);
-		}
-	}
+static u32 ilk_lut_10(const struct drm_color_lut *color)
+{
+	return drm_color_lut_extract(color->red, 10) << 20 |
+		drm_color_lut_extract(color->green, 10) << 10 |
+		drm_color_lut_extract(color->blue, 10);
 }
 
-static void i9xx_load_luts(const struct intel_crtc_state *crtc_state)
+static void ilk_lut_10_pack(struct drm_color_lut *entry, u32 val)
 {
-	i9xx_load_luts_internal(crtc_state, crtc_state->hw.gamma_lut);
+	entry->red = intel_color_lut_pack(REG_FIELD_GET(PREC_PALETTE_RED_MASK, val), 10);
+	entry->green = intel_color_lut_pack(REG_FIELD_GET(PREC_PALETTE_GREEN_MASK, val), 10);
+	entry->blue = intel_color_lut_pack(REG_FIELD_GET(PREC_PALETTE_BLUE_MASK, val), 10);
 }
 
 static void i9xx_color_commit(const struct intel_crtc_state *crtc_state)
@@ -525,6 +526,35 @@ static void skl_color_commit(const struct intel_crtc_state *crtc_state)
 		ilk_load_csc_matrix(crtc_state);
 }
 
+static void i9xx_load_lut_8(struct intel_crtc *crtc,
+			    const struct drm_property_blob *blob)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct drm_color_lut *lut;
+	enum pipe pipe = crtc->pipe;
+	int i;
+
+	if (!blob)
+		return;
+
+	lut = blob->data;
+
+	for (i = 0; i < 256; i++)
+		intel_de_write(dev_priv, PALETTE(pipe, i),
+			       i9xx_lut_8(&lut[i]));
+}
+
+static void i9xx_load_luts(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 drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+
+	assert_pll_enabled(dev_priv, crtc->pipe);
+
+	i9xx_load_lut_8(crtc, gamma_lut);
+}
+
 static void i965_load_lut_10p6(struct intel_crtc *crtc,
 			       const struct drm_property_blob *blob)
 {
@@ -548,14 +578,38 @@ 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->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
 
+	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI))
+		assert_dsi_pll_enabled(dev_priv);
+	else
+		assert_pll_enabled(dev_priv, crtc->pipe);
+
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		i9xx_load_luts(crtc_state);
+		i9xx_load_lut_8(crtc, gamma_lut);
 	else
 		i965_load_lut_10p6(crtc, gamma_lut);
 }
 
+static void ilk_load_lut_8(struct intel_crtc *crtc,
+			   const struct drm_property_blob *blob)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct drm_color_lut *lut;
+	enum pipe pipe = crtc->pipe;
+	int i;
+
+	if (!blob)
+		return;
+
+	lut = blob->data;
+
+	for (i = 0; i < 256; i++)
+		intel_de_write(dev_priv, LGC_PALETTE(pipe, i),
+			       i9xx_lut_8(&lut[i]));
+}
+
 static void ilk_load_lut_10(struct intel_crtc *crtc,
 			    const struct drm_property_blob *blob)
 {
@@ -566,7 +620,7 @@ static void ilk_load_lut_10(struct intel_crtc *crtc,
 
 	for (i = 0; i < lut_size; i++)
 		intel_de_write(dev_priv, PREC_PALETTE(pipe, i),
-		               ilk_lut_10(&lut[i]));
+			       ilk_lut_10(&lut[i]));
 }
 
 static void ilk_load_luts(const struct intel_crtc_state *crtc_state)
@@ -575,7 +629,7 @@ static void ilk_load_luts(const struct intel_crtc_state *crtc_state)
 	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);
+		ilk_load_lut_8(crtc, gamma_lut);
 	else
 		ilk_load_lut_10(crtc, gamma_lut);
 }
@@ -685,7 +739,7 @@ static void ivb_load_luts(const struct intel_crtc_state *crtc_state)
 	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);
+		ilk_load_lut_8(crtc, gamma_lut);
 	} else if (crtc_state->gamma_mode == GAMMA_MODE_MODE_SPLIT) {
 		ivb_load_lut_10(crtc, degamma_lut, PAL_PREC_SPLIT_MODE |
 				PAL_PREC_INDEX_VALUE(0));
@@ -708,7 +762,7 @@ static void bdw_load_luts(const struct intel_crtc_state *crtc_state)
 	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);
+		ilk_load_lut_8(crtc, gamma_lut);
 	} else if (crtc_state->gamma_mode == GAMMA_MODE_MODE_SPLIT) {
 		bdw_load_lut_10(crtc, degamma_lut, PAL_PREC_SPLIT_MODE |
 				PAL_PREC_INDEX_VALUE(0));
@@ -729,9 +783,8 @@ static void glk_load_degamma_lut(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);
 	enum pipe pipe = crtc->pipe;
-	const u32 lut_size = INTEL_INFO(dev_priv)->color.degamma_lut_size;
+	int i, lut_size = INTEL_INFO(dev_priv)->color.degamma_lut_size;
 	const struct drm_color_lut *lut = crtc_state->hw.degamma_lut->data;
-	u32 i;
 
 	/*
 	 * When setting the auto-increment bit, the hardware seems to
@@ -770,8 +823,7 @@ static void glk_load_degamma_lut_linear(const struct intel_crtc_state *crtc_stat
 	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;
-	u32 i;
+	int i, lut_size = INTEL_INFO(dev_priv)->color.degamma_lut_size;
 
 	/*
 	 * When setting the auto-increment bit, the hardware seems to
@@ -812,7 +864,7 @@ static void glk_load_luts(const struct intel_crtc_state *crtc_state)
 		glk_load_degamma_lut_linear(crtc_state);
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT) {
-		i9xx_load_luts(crtc_state);
+		ilk_load_lut_8(crtc, gamma_lut);
 	} else {
 		bdw_load_lut_10(crtc, gamma_lut, PAL_PREC_INDEX_VALUE(0));
 		ivb_load_lut_ext_max(crtc);
@@ -856,7 +908,7 @@ icl_program_gamma_superfine_segment(const struct intel_crtc_state *crtc_state)
 	const struct drm_color_lut *lut = blob->data;
 	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
-	u32 i;
+	int i;
 
 	/*
 	 * Program Super Fine segment (let's call it seg1)...
@@ -889,7 +941,7 @@ icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
 	const struct drm_color_lut *entry;
 	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
-	u32 i;
+	int i;
 
 	/*
 	 * Program Fine segment (let's call it seg2)...
@@ -948,7 +1000,7 @@ static void icl_load_luts(const struct intel_crtc_state *crtc_state)
 
 	switch (crtc_state->gamma_mode & GAMMA_MODE_MODE_MASK) {
 	case GAMMA_MODE_MODE_8BIT:
-		i9xx_load_luts(crtc_state);
+		ilk_load_lut_8(crtc, gamma_lut);
 		break;
 	case GAMMA_MODE_MODE_12BIT_MULTI_SEGMENTED:
 		icl_program_gamma_superfine_segment(crtc_state);
@@ -974,6 +1026,13 @@ static u32 chv_cgm_degamma_udw(const struct drm_color_lut *color)
 	return drm_color_lut_extract(color->red, 14);
 }
 
+static void chv_cgm_gamma_pack(struct drm_color_lut *entry, u32 ldw, u32 udw)
+{
+	entry->green = intel_color_lut_pack(REG_FIELD_GET(CGM_PIPE_GAMMA_GREEN_MASK, ldw), 10);
+	entry->blue = intel_color_lut_pack(REG_FIELD_GET(CGM_PIPE_GAMMA_BLUE_MASK, ldw), 10);
+	entry->red = intel_color_lut_pack(REG_FIELD_GET(CGM_PIPE_GAMMA_RED_MASK, udw), 10);
+}
+
 static void chv_load_cgm_degamma(struct intel_crtc *crtc,
 				 const struct drm_property_blob *blob)
 {
@@ -1020,21 +1079,24 @@ 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->uapi.crtc);
-	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	const struct drm_property_blob *degamma_lut = crtc_state->hw.degamma_lut;
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	const struct drm_property_blob *ctm = crtc_state->hw.ctm;
 
-	cherryview_load_csc_matrix(crtc_state);
+	if (crtc_state->cgm_mode & CGM_PIPE_MODE_CSC)
+		chv_load_cgm_csc(crtc, ctm);
 
-	if (crtc_state_is_legacy_gamma(crtc_state)) {
-		i9xx_load_luts(crtc_state);
-		return;
-	}
-
-	if (degamma_lut)
+	if (crtc_state->cgm_mode & CGM_PIPE_MODE_DEGAMMA)
 		chv_load_cgm_degamma(crtc, degamma_lut);
 
-	if (gamma_lut)
+	if (crtc_state->cgm_mode & CGM_PIPE_MODE_GAMMA)
 		chv_load_cgm_gamma(crtc, gamma_lut);
+	else
+		i965_load_luts(crtc_state);
+
+	intel_de_write(dev_priv, CGM_PIPE_MODE(crtc->pipe),
+		       crtc_state->cgm_mode);
 }
 
 void intel_color_load_luts(const struct intel_crtc_state *crtc_state)
@@ -1660,28 +1722,13 @@ bool intel_color_lut_equal(struct drm_property_blob *blob1,
 	return true;
 }
 
-/* convert hw value with given bit_precision to lut property val */
-static u32 intel_color_lut_pack(u32 val, u32 bit_precision)
+static struct drm_property_blob *i9xx_read_lut_8(struct intel_crtc *crtc)
 {
-	u32 max = 0xffff >> (16 - bit_precision);
-
-	val = clamp_val(val, 0, max);
-
-	if (bit_precision < 16)
-		val <<= 16 - bit_precision;
-
-	return val;
-}
-
-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->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	struct drm_property_blob *blob;
-	struct drm_color_lut *blob_data;
-	u32 i, val;
+	struct drm_color_lut *lut;
+	int i;
 
 	blob = drm_property_create_blob(&dev_priv->drm,
 					sizeof(struct drm_color_lut) * LEGACY_LUT_LENGTH,
@@ -1689,20 +1736,12 @@ i9xx_read_lut_8(const struct intel_crtc_state *crtc_state)
 	if (IS_ERR(blob))
 		return NULL;
 
-	blob_data = blob->data;
+	lut = blob->data;
 
 	for (i = 0; i < LEGACY_LUT_LENGTH; i++) {
-		if (HAS_GMCH(dev_priv))
-			val = intel_de_read(dev_priv, PALETTE(pipe, i));
-		else
-			val = intel_de_read(dev_priv, LGC_PALETTE(pipe, i));
-
-		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
-							LGC_PALETTE_RED_MASK, val), 8);
-		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
-							  LGC_PALETTE_GREEN_MASK, val), 8);
-		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
-							 LGC_PALETTE_BLUE_MASK, val), 8);
+		u32 val = intel_de_read(dev_priv, PALETTE(pipe, i));
+
+		i9xx_lut_8_pack(&lut[i], val);
 	}
 
 	return blob;
@@ -1710,22 +1749,21 @@ i9xx_read_lut_8(const struct intel_crtc_state *crtc_state)
 
 static void i9xx_read_luts(struct intel_crtc_state *crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+
 	if (!crtc_state->gamma_enable)
 		return;
 
-	crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
+	crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc);
 }
 
-static struct drm_property_blob *
-i965_read_lut_10p6(const struct intel_crtc_state *crtc_state)
+static struct drm_property_blob *i965_read_lut_10p6(struct intel_crtc *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;
+	int i, lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 	enum pipe pipe = crtc->pipe;
 	struct drm_property_blob *blob;
-	struct drm_color_lut *blob_data;
-	u32 i, val1, val2;
+	struct drm_color_lut *lut;
 
 	blob = drm_property_create_blob(&dev_priv->drm,
 					sizeof(struct drm_color_lut) * lut_size,
@@ -1733,51 +1771,42 @@ i965_read_lut_10p6(const struct intel_crtc_state *crtc_state)
 	if (IS_ERR(blob))
 		return NULL;
 
-	blob_data = blob->data;
+	lut = blob->data;
 
 	for (i = 0; i < lut_size - 1; i++) {
-		val1 = intel_de_read(dev_priv, PALETTE(pipe, 2 * i + 0));
-		val2 = intel_de_read(dev_priv, PALETTE(pipe, 2 * i + 1));
-
-		blob_data[i].red = REG_FIELD_GET(PALETTE_RED_MASK, val2) << 8 |
-						 REG_FIELD_GET(PALETTE_RED_MASK, val1);
-		blob_data[i].green = REG_FIELD_GET(PALETTE_GREEN_MASK, val2) << 8 |
-						   REG_FIELD_GET(PALETTE_GREEN_MASK, val1);
-		blob_data[i].blue = REG_FIELD_GET(PALETTE_BLUE_MASK, val2) << 8 |
-						  REG_FIELD_GET(PALETTE_BLUE_MASK, val1);
+		u32 ldw = intel_de_read(dev_priv, PALETTE(pipe, 2 * i + 0));
+		u32 udw = intel_de_read(dev_priv, PALETTE(pipe, 2 * i + 1));
+
+		i965_lut_10p6_pack(&lut[i], ldw, udw);
 	}
 
-	blob_data[i].red = REG_FIELD_GET(PIPEGCMAX_RGB_MASK,
-					 intel_de_read(dev_priv, PIPEGCMAX(pipe, 0)));
-	blob_data[i].green = REG_FIELD_GET(PIPEGCMAX_RGB_MASK,
-					   intel_de_read(dev_priv, PIPEGCMAX(pipe, 1)));
-	blob_data[i].blue = REG_FIELD_GET(PIPEGCMAX_RGB_MASK,
-					  intel_de_read(dev_priv, PIPEGCMAX(pipe, 2)));
+	lut[i].red = i965_lut_11p6_max_pack(intel_de_read(dev_priv, PIPEGCMAX(pipe, 0)));
+	lut[i].green = i965_lut_11p6_max_pack(intel_de_read(dev_priv, PIPEGCMAX(pipe, 1)));
+	lut[i].blue = i965_lut_11p6_max_pack(intel_de_read(dev_priv, PIPEGCMAX(pipe, 2)));
 
 	return blob;
 }
 
 static void i965_read_luts(struct intel_crtc_state *crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+
 	if (!crtc_state->gamma_enable)
 		return;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
+		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc);
 	else
-		crtc_state->hw.gamma_lut = i965_read_lut_10p6(crtc_state);
+		crtc_state->hw.gamma_lut = i965_read_lut_10p6(crtc);
 }
 
-static struct drm_property_blob *
-chv_read_cgm_lut(const struct intel_crtc_state *crtc_state)
+static struct drm_property_blob *chv_read_cgm_gamma(struct intel_crtc *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;
+	int i, lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 	enum pipe pipe = crtc->pipe;
 	struct drm_property_blob *blob;
-	struct drm_color_lut *blob_data;
-	u32 i, val;
+	struct drm_color_lut *lut;
 
 	blob = drm_property_create_blob(&dev_priv->drm,
 					sizeof(struct drm_color_lut) * lut_size,
@@ -1785,18 +1814,13 @@ chv_read_cgm_lut(const struct intel_crtc_state *crtc_state)
 	if (IS_ERR(blob))
 		return NULL;
 
-	blob_data = blob->data;
+	lut = blob->data;
 
 	for (i = 0; i < lut_size; i++) {
-		val = intel_de_read(dev_priv, CGM_PIPE_GAMMA(pipe, i, 0));
-		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
-							  CGM_PIPE_GAMMA_GREEN_MASK, val), 10);
-		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
-							 CGM_PIPE_GAMMA_BLUE_MASK, val), 10);
-
-		val = intel_de_read(dev_priv, CGM_PIPE_GAMMA(pipe, i, 1));
-		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
-							CGM_PIPE_GAMMA_RED_MASK, val), 10);
+		u32 ldw = intel_de_read(dev_priv, CGM_PIPE_GAMMA(pipe, i, 0));
+		u32 udw = intel_de_read(dev_priv, CGM_PIPE_GAMMA(pipe, i, 1));
+
+		chv_cgm_gamma_pack(&lut[i], ldw, udw);
 	}
 
 	return blob;
@@ -1804,22 +1828,46 @@ chv_read_cgm_lut(const struct intel_crtc_state *crtc_state)
 
 static void chv_read_luts(struct intel_crtc_state *crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+
 	if (crtc_state->cgm_mode & CGM_PIPE_MODE_GAMMA)
-		crtc_state->hw.gamma_lut = chv_read_cgm_lut(crtc_state);
+		crtc_state->hw.gamma_lut = chv_read_cgm_gamma(crtc);
 	else
 		i965_read_luts(crtc_state);
 }
 
-static struct drm_property_blob *
-ilk_read_lut_10(const struct intel_crtc_state *crtc_state)
+static struct drm_property_blob *ilk_read_lut_8(struct intel_crtc *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;
 	struct drm_property_blob *blob;
-	struct drm_color_lut *blob_data;
-	u32 i, val;
+	struct drm_color_lut *lut;
+	int i;
+
+	blob = drm_property_create_blob(&dev_priv->drm,
+					sizeof(struct drm_color_lut) * LEGACY_LUT_LENGTH,
+					NULL);
+	if (IS_ERR(blob))
+		return NULL;
+
+	lut = blob->data;
+
+	for (i = 0; i < LEGACY_LUT_LENGTH; i++) {
+		u32 val = intel_de_read(dev_priv, LGC_PALETTE(pipe, i));
+
+		i9xx_lut_8_pack(&lut[i], val);
+	}
+
+	return blob;
+}
+
+static struct drm_property_blob *ilk_read_lut_10(struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	int i, lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
+	enum pipe pipe = crtc->pipe;
+	struct drm_property_blob *blob;
+	struct drm_color_lut *lut;
 
 	blob = drm_property_create_blob(&dev_priv->drm,
 					sizeof(struct drm_color_lut) * lut_size,
@@ -1827,17 +1875,12 @@ ilk_read_lut_10(const struct intel_crtc_state *crtc_state)
 	if (IS_ERR(blob))
 		return NULL;
 
-	blob_data = blob->data;
+	lut = blob->data;
 
 	for (i = 0; i < lut_size; i++) {
-		val = intel_de_read(dev_priv, PREC_PALETTE(pipe, i));
-
-		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
-							PREC_PALETTE_RED_MASK, val), 10);
-		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
-							  PREC_PALETTE_GREEN_MASK, val), 10);
-		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
-							 PREC_PALETTE_BLUE_MASK, val), 10);
+		u32 val = intel_de_read(dev_priv, PREC_PALETTE(pipe, i));
+
+		ilk_lut_10_pack(&lut[i], val);
 	}
 
 	return blob;
@@ -1845,6 +1888,8 @@ ilk_read_lut_10(const struct intel_crtc_state *crtc_state)
 
 static void ilk_read_luts(struct intel_crtc_state *crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+
 	if (!crtc_state->gamma_enable)
 		return;
 
@@ -1852,21 +1897,19 @@ static void ilk_read_luts(struct intel_crtc_state *crtc_state)
 		return;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
+		crtc_state->hw.gamma_lut = ilk_read_lut_8(crtc);
 	else
-		crtc_state->hw.gamma_lut = ilk_read_lut_10(crtc_state);
+		crtc_state->hw.gamma_lut = ilk_read_lut_10(crtc);
 }
 
-static struct drm_property_blob *
-glk_read_lut_10(const struct intel_crtc_state *crtc_state, u32 prec_index)
+static struct drm_property_blob *glk_read_lut_10(struct intel_crtc *crtc,
+						 u32 prec_index)
 {
-	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);
+	int i, hw_lut_size = ivb_lut_10_size(prec_index);
 	enum pipe pipe = crtc->pipe;
 	struct drm_property_blob *blob;
-	struct drm_color_lut *blob_data;
-	u32 i, val;
+	struct drm_color_lut *lut;
 
 	blob = drm_property_create_blob(&dev_priv->drm,
 					sizeof(struct drm_color_lut) * hw_lut_size,
@@ -1874,20 +1917,15 @@ glk_read_lut_10(const struct intel_crtc_state *crtc_state, u32 prec_index)
 	if (IS_ERR(blob))
 		return NULL;
 
-	blob_data = blob->data;
+	lut = blob->data;
 
 	intel_de_write(dev_priv, PREC_PAL_INDEX(pipe),
 		       prec_index | PAL_PREC_AUTO_INCREMENT);
 
 	for (i = 0; i < hw_lut_size; i++) {
-		val = intel_de_read(dev_priv, PREC_PAL_DATA(pipe));
-
-		blob_data[i].red = intel_color_lut_pack(REG_FIELD_GET(
-							PREC_PAL_DATA_RED_MASK, val), 10);
-		blob_data[i].green = intel_color_lut_pack(REG_FIELD_GET(
-							PREC_PAL_DATA_GREEN_MASK, val), 10);
-		blob_data[i].blue = intel_color_lut_pack(REG_FIELD_GET(
-							PREC_PAL_DATA_BLUE_MASK, val), 10);
+		u32 val = intel_de_read(dev_priv, PREC_PAL_DATA(pipe));
+
+		ilk_lut_10_pack(&lut[i], val);
 	}
 
 	intel_de_write(dev_priv, PREC_PAL_INDEX(pipe), 0);
@@ -1897,13 +1935,15 @@ glk_read_lut_10(const struct intel_crtc_state *crtc_state, u32 prec_index)
 
 static void glk_read_luts(struct intel_crtc_state *crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+
 	if (!crtc_state->gamma_enable)
 		return;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
+		crtc_state->hw.gamma_lut = ilk_read_lut_8(crtc);
 	else
-		crtc_state->hw.gamma_lut = glk_read_lut_10(crtc_state, PAL_PREC_INDEX_VALUE(0));
+		crtc_state->hw.gamma_lut = glk_read_lut_10(crtc, 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 f49c98f6cb7e..78f9b6cde810 100644
--- a/drivers/gpu/drm/i915/display/intel_crt.c
+++ b/drivers/gpu/drm/i915/display/intel_crt.c
@@ -32,7 +32,6 @@
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_probe_helper.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_connector.h"
diff --git a/drivers/gpu/drm/i915/display/intel_csr.c b/drivers/gpu/drm/i915/display/intel_csr.c
index 57320c12839f..3112572cfb7d 100644
--- a/drivers/gpu/drm/i915/display/intel_csr.c
+++ b/drivers/gpu/drm/i915/display/intel_csr.c
@@ -40,8 +40,8 @@
 
 #define GEN12_CSR_MAX_FW_SIZE		ICL_CSR_MAX_FW_SIZE
 
-#define TGL_CSR_PATH			"i915/tgl_dmc_ver2_04.bin"
-#define TGL_CSR_VERSION_REQUIRED	CSR_VERSION(2, 4)
+#define TGL_CSR_PATH			"i915/tgl_dmc_ver2_06.bin"
+#define TGL_CSR_VERSION_REQUIRED	CSR_VERSION(2, 6)
 #define TGL_CSR_MAX_FW_SIZE		0x6000
 MODULE_FIRMWARE(TGL_CSR_PATH);
 
diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c
index 9f7d1d7189ae..73d0f4648c06 100644
--- a/drivers/gpu/drm/i915/display/intel_ddi.c
+++ b/drivers/gpu/drm/i915/display/intel_ddi.c
@@ -1325,164 +1325,6 @@ intel_ddi_get_crtc_encoder(struct intel_crtc *crtc)
 	return ret;
 }
 
-static int hsw_ddi_calc_wrpll_link(struct drm_i915_private *dev_priv,
-				   i915_reg_t reg)
-{
-	int refclk;
-	int n, p, r;
-	u32 wrpll;
-
-	wrpll = intel_de_read(dev_priv, reg);
-	switch (wrpll & WRPLL_REF_MASK) {
-	case WRPLL_REF_SPECIAL_HSW:
-		/*
-		 * muxed-SSC for BDW.
-		 * non-SSC for non-ULT HSW. Check FUSE_STRAP3
-		 * for the non-SSC reference frequency.
-		 */
-		if (IS_HASWELL(dev_priv) && !IS_HSW_ULT(dev_priv)) {
-			if (intel_de_read(dev_priv, FUSE_STRAP3) & HSW_REF_CLK_SELECT)
-				refclk = 24;
-			else
-				refclk = 135;
-			break;
-		}
-		/* fall through */
-	case WRPLL_REF_PCH_SSC:
-		/*
-		 * We could calculate spread here, but our checking
-		 * code only cares about 5% accuracy, and spread is a max of
-		 * 0.5% downspread.
-		 */
-		refclk = 135;
-		break;
-	case WRPLL_REF_LCPLL:
-		refclk = 2700;
-		break;
-	default:
-		MISSING_CASE(wrpll);
-		return 0;
-	}
-
-	r = wrpll & WRPLL_DIVIDER_REF_MASK;
-	p = (wrpll & WRPLL_DIVIDER_POST_MASK) >> WRPLL_DIVIDER_POST_SHIFT;
-	n = (wrpll & WRPLL_DIVIDER_FB_MASK) >> WRPLL_DIVIDER_FB_SHIFT;
-
-	/* Convert to KHz, p & r have a fixed point portion */
-	return (refclk * n * 100) / (p * r);
-}
-
-static int skl_calc_wrpll_link(const struct intel_dpll_hw_state *pll_state)
-{
-	u32 p0, p1, p2, dco_freq;
-
-	p0 = pll_state->cfgcr2 & DPLL_CFGCR2_PDIV_MASK;
-	p2 = pll_state->cfgcr2 & DPLL_CFGCR2_KDIV_MASK;
-
-	if (pll_state->cfgcr2 &  DPLL_CFGCR2_QDIV_MODE(1))
-		p1 = (pll_state->cfgcr2 & DPLL_CFGCR2_QDIV_RATIO_MASK) >> 8;
-	else
-		p1 = 1;
-
-
-	switch (p0) {
-	case DPLL_CFGCR2_PDIV_1:
-		p0 = 1;
-		break;
-	case DPLL_CFGCR2_PDIV_2:
-		p0 = 2;
-		break;
-	case DPLL_CFGCR2_PDIV_3:
-		p0 = 3;
-		break;
-	case DPLL_CFGCR2_PDIV_7:
-		p0 = 7;
-		break;
-	}
-
-	switch (p2) {
-	case DPLL_CFGCR2_KDIV_5:
-		p2 = 5;
-		break;
-	case DPLL_CFGCR2_KDIV_2:
-		p2 = 2;
-		break;
-	case DPLL_CFGCR2_KDIV_3:
-		p2 = 3;
-		break;
-	case DPLL_CFGCR2_KDIV_1:
-		p2 = 1;
-		break;
-	}
-
-	dco_freq = (pll_state->cfgcr1 & DPLL_CFGCR1_DCO_INTEGER_MASK)
-		* 24 * 1000;
-
-	dco_freq += (((pll_state->cfgcr1 & DPLL_CFGCR1_DCO_FRACTION_MASK) >> 9)
-		     * 24 * 1000) / 0x8000;
-
-	if (WARN_ON(p0 == 0 || p1 == 0 || p2 == 0))
-		return 0;
-
-	return dco_freq / (p0 * p1 * p2 * 5);
-}
-
-int cnl_calc_wrpll_link(struct drm_i915_private *dev_priv,
-			struct intel_dpll_hw_state *pll_state)
-{
-	u32 p0, p1, p2, dco_freq, ref_clock;
-
-	p0 = pll_state->cfgcr1 & DPLL_CFGCR1_PDIV_MASK;
-	p2 = pll_state->cfgcr1 & DPLL_CFGCR1_KDIV_MASK;
-
-	if (pll_state->cfgcr1 & DPLL_CFGCR1_QDIV_MODE(1))
-		p1 = (pll_state->cfgcr1 & DPLL_CFGCR1_QDIV_RATIO_MASK) >>
-			DPLL_CFGCR1_QDIV_RATIO_SHIFT;
-	else
-		p1 = 1;
-
-
-	switch (p0) {
-	case DPLL_CFGCR1_PDIV_2:
-		p0 = 2;
-		break;
-	case DPLL_CFGCR1_PDIV_3:
-		p0 = 3;
-		break;
-	case DPLL_CFGCR1_PDIV_5:
-		p0 = 5;
-		break;
-	case DPLL_CFGCR1_PDIV_7:
-		p0 = 7;
-		break;
-	}
-
-	switch (p2) {
-	case DPLL_CFGCR1_KDIV_1:
-		p2 = 1;
-		break;
-	case DPLL_CFGCR1_KDIV_2:
-		p2 = 2;
-		break;
-	case DPLL_CFGCR1_KDIV_3:
-		p2 = 3;
-		break;
-	}
-
-	ref_clock = cnl_hdmi_pll_ref_clock(dev_priv);
-
-	dco_freq = (pll_state->cfgcr0 & DPLL_CFGCR0_DCO_INTEGER_MASK)
-		* ref_clock;
-
-	dco_freq += (((pll_state->cfgcr0 & DPLL_CFGCR0_DCO_FRACTION_MASK) >>
-		      DPLL_CFGCR0_DCO_FRACTION_SHIFT) * ref_clock) / 0x8000;
-
-	if (drm_WARN_ON(&dev_priv->drm, p0 == 0 || p1 == 0 || p2 == 0))
-		return 0;
-
-	return dco_freq / (p0 * p1 * p2 * 5);
-}
-
 static int icl_calc_tbt_pll_link(struct drm_i915_private *dev_priv,
 				 enum port port)
 {
@@ -1505,77 +1347,6 @@ static int icl_calc_tbt_pll_link(struct drm_i915_private *dev_priv,
 	}
 }
 
-static int icl_calc_mg_pll_link(struct drm_i915_private *dev_priv,
-				const struct intel_dpll_hw_state *pll_state)
-{
-	u32 m1, m2_int, m2_frac, div1, div2, ref_clock;
-	u64 tmp;
-
-	ref_clock = dev_priv->cdclk.hw.ref;
-
-	if (INTEL_GEN(dev_priv) >= 12) {
-		m1 = pll_state->mg_pll_div0 & DKL_PLL_DIV0_FBPREDIV_MASK;
-		m1 = m1 >> DKL_PLL_DIV0_FBPREDIV_SHIFT;
-		m2_int = pll_state->mg_pll_div0 & DKL_PLL_DIV0_FBDIV_INT_MASK;
-
-		if (pll_state->mg_pll_bias & DKL_PLL_BIAS_FRAC_EN_H) {
-			m2_frac = pll_state->mg_pll_bias &
-				  DKL_PLL_BIAS_FBDIV_FRAC_MASK;
-			m2_frac = m2_frac >> DKL_PLL_BIAS_FBDIV_SHIFT;
-		} else {
-			m2_frac = 0;
-		}
-	} else {
-		m1 = pll_state->mg_pll_div1 & MG_PLL_DIV1_FBPREDIV_MASK;
-		m2_int = pll_state->mg_pll_div0 & MG_PLL_DIV0_FBDIV_INT_MASK;
-
-		if (pll_state->mg_pll_div0 & MG_PLL_DIV0_FRACNEN_H) {
-			m2_frac = pll_state->mg_pll_div0 &
-				  MG_PLL_DIV0_FBDIV_FRAC_MASK;
-			m2_frac = m2_frac >> MG_PLL_DIV0_FBDIV_FRAC_SHIFT;
-		} else {
-			m2_frac = 0;
-		}
-	}
-
-	switch (pll_state->mg_clktop2_hsclkctl &
-		MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_MASK) {
-	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_2:
-		div1 = 2;
-		break;
-	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_3:
-		div1 = 3;
-		break;
-	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_5:
-		div1 = 5;
-		break;
-	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_7:
-		div1 = 7;
-		break;
-	default:
-		MISSING_CASE(pll_state->mg_clktop2_hsclkctl);
-		return 0;
-	}
-
-	div2 = (pll_state->mg_clktop2_hsclkctl &
-		MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO_MASK) >>
-		MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO_SHIFT;
-
-	/* div2 value of 0 is same as 1 means no div */
-	if (div2 == 0)
-		div2 = 1;
-
-	/*
-	 * Adjust the original formula to delay the division by 2^22 in order to
-	 * minimize possible rounding errors.
-	 */
-	tmp = (u64)m1 * m2_int * ref_clock +
-	      (((u64)m1 * m2_frac * ref_clock) >> 22);
-	tmp = div_u64(tmp, 5 * div1 * div2);
-
-	return tmp;
-}
-
 static void ddi_dotclock_get(struct intel_crtc_state *pipe_config)
 {
 	int dotclock;
@@ -1601,215 +1372,22 @@ static void ddi_dotclock_get(struct intel_crtc_state *pipe_config)
 	pipe_config->hw.adjusted_mode.crtc_clock = dotclock;
 }
 
-static void icl_ddi_clock_get(struct intel_encoder *encoder,
-			      struct intel_crtc_state *pipe_config)
-{
-	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_dpll_hw_state *pll_state = &pipe_config->dpll_hw_state;
-	enum port port = encoder->port;
-	enum phy phy = intel_port_to_phy(dev_priv, port);
-	int link_clock;
-
-	if (intel_phy_is_combo(dev_priv, phy)) {
-		link_clock = cnl_calc_wrpll_link(dev_priv, pll_state);
-	} else {
-		enum intel_dpll_id pll_id = intel_get_shared_dpll_id(dev_priv,
-						pipe_config->shared_dpll);
-
-		if (pll_id == DPLL_ID_ICL_TBTPLL)
-			link_clock = icl_calc_tbt_pll_link(dev_priv, port);
-		else
-			link_clock = icl_calc_mg_pll_link(dev_priv, pll_state);
-	}
-
-	pipe_config->port_clock = link_clock;
-
-	ddi_dotclock_get(pipe_config);
-}
-
-static void cnl_ddi_clock_get(struct intel_encoder *encoder,
-			      struct intel_crtc_state *pipe_config)
-{
-	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_dpll_hw_state *pll_state = &pipe_config->dpll_hw_state;
-	int link_clock;
-
-	if (pll_state->cfgcr0 & DPLL_CFGCR0_HDMI_MODE) {
-		link_clock = cnl_calc_wrpll_link(dev_priv, pll_state);
-	} else {
-		link_clock = pll_state->cfgcr0 & DPLL_CFGCR0_LINK_RATE_MASK;
-
-		switch (link_clock) {
-		case DPLL_CFGCR0_LINK_RATE_810:
-			link_clock = 81000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_1080:
-			link_clock = 108000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_1350:
-			link_clock = 135000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_1620:
-			link_clock = 162000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_2160:
-			link_clock = 216000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_2700:
-			link_clock = 270000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_3240:
-			link_clock = 324000;
-			break;
-		case DPLL_CFGCR0_LINK_RATE_4050:
-			link_clock = 405000;
-			break;
-		default:
-			drm_WARN(&dev_priv->drm, 1, "Unsupported link rate\n");
-			break;
-		}
-		link_clock *= 2;
-	}
-
-	pipe_config->port_clock = link_clock;
-
-	ddi_dotclock_get(pipe_config);
-}
-
-static void skl_ddi_clock_get(struct intel_encoder *encoder,
-			      struct intel_crtc_state *pipe_config)
-{
-	struct intel_dpll_hw_state *pll_state = &pipe_config->dpll_hw_state;
-	int link_clock;
-
-	/*
-	 * ctrl1 register is already shifted for each pll, just use 0 to get
-	 * the internal shift for each field
-	 */
-	if (pll_state->ctrl1 & DPLL_CTRL1_HDMI_MODE(0)) {
-		link_clock = skl_calc_wrpll_link(pll_state);
-	} else {
-		link_clock = pll_state->ctrl1 & DPLL_CTRL1_LINK_RATE_MASK(0);
-		link_clock >>= DPLL_CTRL1_LINK_RATE_SHIFT(0);
-
-		switch (link_clock) {
-		case DPLL_CTRL1_LINK_RATE_810:
-			link_clock = 81000;
-			break;
-		case DPLL_CTRL1_LINK_RATE_1080:
-			link_clock = 108000;
-			break;
-		case DPLL_CTRL1_LINK_RATE_1350:
-			link_clock = 135000;
-			break;
-		case DPLL_CTRL1_LINK_RATE_1620:
-			link_clock = 162000;
-			break;
-		case DPLL_CTRL1_LINK_RATE_2160:
-			link_clock = 216000;
-			break;
-		case DPLL_CTRL1_LINK_RATE_2700:
-			link_clock = 270000;
-			break;
-		default:
-			drm_WARN(encoder->base.dev, 1,
-				 "Unsupported link rate\n");
-			break;
-		}
-		link_clock *= 2;
-	}
-
-	pipe_config->port_clock = link_clock;
-
-	ddi_dotclock_get(pipe_config);
-}
-
-static void hsw_ddi_clock_get(struct intel_encoder *encoder,
-			      struct intel_crtc_state *pipe_config)
-{
-	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	int link_clock = 0;
-	u32 val, pll;
-
-	val = hsw_pll_to_ddi_pll_sel(pipe_config->shared_dpll);
-	switch (val & PORT_CLK_SEL_MASK) {
-	case PORT_CLK_SEL_LCPLL_810:
-		link_clock = 81000;
-		break;
-	case PORT_CLK_SEL_LCPLL_1350:
-		link_clock = 135000;
-		break;
-	case PORT_CLK_SEL_LCPLL_2700:
-		link_clock = 270000;
-		break;
-	case PORT_CLK_SEL_WRPLL1:
-		link_clock = hsw_ddi_calc_wrpll_link(dev_priv, WRPLL_CTL(0));
-		break;
-	case PORT_CLK_SEL_WRPLL2:
-		link_clock = hsw_ddi_calc_wrpll_link(dev_priv, WRPLL_CTL(1));
-		break;
-	case PORT_CLK_SEL_SPLL:
-		pll = intel_de_read(dev_priv, SPLL_CTL) & SPLL_FREQ_MASK;
-		if (pll == SPLL_FREQ_810MHz)
-			link_clock = 81000;
-		else if (pll == SPLL_FREQ_1350MHz)
-			link_clock = 135000;
-		else if (pll == SPLL_FREQ_2700MHz)
-			link_clock = 270000;
-		else {
-			drm_WARN(&dev_priv->drm, 1, "bad spll freq\n");
-			return;
-		}
-		break;
-	default:
-		drm_WARN(&dev_priv->drm, 1, "bad port clock sel\n");
-		return;
-	}
-
-	pipe_config->port_clock = link_clock * 2;
-
-	ddi_dotclock_get(pipe_config);
-}
-
-static int bxt_calc_pll_link(const struct intel_dpll_hw_state *pll_state)
-{
-	struct dpll clock;
-
-	clock.m1 = 2;
-	clock.m2 = (pll_state->pll0 & PORT_PLL_M2_MASK) << 22;
-	if (pll_state->pll3 & PORT_PLL_M2_FRAC_ENABLE)
-		clock.m2 |= pll_state->pll2 & PORT_PLL_M2_FRAC_MASK;
-	clock.n = (pll_state->pll1 & PORT_PLL_N_MASK) >> PORT_PLL_N_SHIFT;
-	clock.p1 = (pll_state->ebb0 & PORT_PLL_P1_MASK) >> PORT_PLL_P1_SHIFT;
-	clock.p2 = (pll_state->ebb0 & PORT_PLL_P2_MASK) >> PORT_PLL_P2_SHIFT;
-
-	return chv_calc_dpll_params(100000, &clock);
-}
-
-static void bxt_ddi_clock_get(struct intel_encoder *encoder,
-			      struct intel_crtc_state *pipe_config)
-{
-	pipe_config->port_clock =
-		bxt_calc_pll_link(&pipe_config->dpll_hw_state);
-
-	ddi_dotclock_get(pipe_config);
-}
-
 static void intel_ddi_clock_get(struct intel_encoder *encoder,
 				struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
 
-	if (INTEL_GEN(dev_priv) >= 11)
-		icl_ddi_clock_get(encoder, pipe_config);
-	else if (IS_CANNONLAKE(dev_priv))
-		cnl_ddi_clock_get(encoder, pipe_config);
-	else if (IS_GEN9_LP(dev_priv))
-		bxt_ddi_clock_get(encoder, pipe_config);
-	else if (IS_GEN9_BC(dev_priv))
-		skl_ddi_clock_get(encoder, pipe_config);
-	else if (INTEL_GEN(dev_priv) <= 8)
-		hsw_ddi_clock_get(encoder, pipe_config);
+	if (intel_phy_is_tc(dev_priv, phy) &&
+	    intel_get_shared_dpll_id(dev_priv, pipe_config->shared_dpll) ==
+	    DPLL_ID_ICL_TBTPLL)
+		pipe_config->port_clock = icl_calc_tbt_pll_link(dev_priv,
+								encoder->port);
+	else
+		pipe_config->port_clock =
+			intel_dpll_get_freq(dev_priv, pipe_config->shared_dpll);
+
+	ddi_dotclock_get(pipe_config);
 }
 
 void intel_ddi_set_dp_msa(const struct intel_crtc_state *crtc_state,
@@ -3049,7 +2627,7 @@ static void icl_map_plls_to_ports(struct intel_encoder *encoder,
 	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
 	u32 val;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 
 	val = intel_de_read(dev_priv, ICL_DPCLKA_CFGCR0);
 	drm_WARN_ON(&dev_priv->drm,
@@ -3075,7 +2653,7 @@ static void icl_map_plls_to_ports(struct intel_encoder *encoder,
 	val &= ~icl_dpclka_cfgcr0_clk_off(dev_priv, phy);
 	intel_de_write(dev_priv, ICL_DPCLKA_CFGCR0, val);
 
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static void icl_unmap_plls_to_ports(struct intel_encoder *encoder)
@@ -3084,13 +2662,13 @@ static void icl_unmap_plls_to_ports(struct intel_encoder *encoder)
 	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
 	u32 val;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 
 	val = intel_de_read(dev_priv, ICL_DPCLKA_CFGCR0);
 	val |= icl_dpclka_cfgcr0_clk_off(dev_priv, phy);
 	intel_de_write(dev_priv, ICL_DPCLKA_CFGCR0, val);
 
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static void icl_sanitize_port_clk_off(struct drm_i915_private *dev_priv,
@@ -3189,7 +2767,7 @@ static void intel_ddi_clk_select(struct intel_encoder *encoder,
 	if (drm_WARN_ON(&dev_priv->drm, !pll))
 		return;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 
 	if (INTEL_GEN(dev_priv) >= 11) {
 		if (!intel_phy_is_combo(dev_priv, phy))
@@ -3233,7 +2811,7 @@ static void intel_ddi_clk_select(struct intel_encoder *encoder,
 			       hsw_pll_to_ddi_pll_sel(pll));
 	}
 
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static void intel_ddi_clk_disable(struct intel_encoder *encoder)
@@ -3987,8 +3565,9 @@ static void intel_enable_ddi_hdmi(struct intel_encoder *encoder,
 	if (!intel_hdmi_handle_sink_scrambling(encoder, connector,
 					       crtc_state->hdmi_high_tmds_clock_ratio,
 					       crtc_state->hdmi_scrambling))
-		DRM_ERROR("[CONNECTOR:%d:%s] Failed to configure sink scrambling/TMDS bit clock ratio\n",
-			  connector->base.id, connector->name);
+		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] Failed to configure sink "
+			      "scrambling/TMDS bit clock ratio\n",
+			       connector->base.id, connector->name);
 
 	/* Display WA #1143: skl,kbl,cfl */
 	if (IS_GEN9_BC(dev_priv)) {
diff --git a/drivers/gpu/drm/i915/display/intel_ddi.h b/drivers/gpu/drm/i915/display/intel_ddi.h
index 167c6579d972..55fd72b901fe 100644
--- a/drivers/gpu/drm/i915/display/intel_ddi.h
+++ b/drivers/gpu/drm/i915/display/intel_ddi.h
@@ -6,8 +6,6 @@
 #ifndef __INTEL_DDI_H__
 #define __INTEL_DDI_H__
 
-#include <drm/i915_drm.h>
-
 #include "intel_display.h"
 
 struct drm_connector_state;
@@ -47,7 +45,5 @@ u8 intel_ddi_dp_pre_emphasis_max(struct intel_encoder *encoder,
 int intel_ddi_toggle_hdcp_signalling(struct intel_encoder *intel_encoder,
 				     bool enable);
 void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder);
-int cnl_calc_wrpll_link(struct drm_i915_private *dev_priv,
-			struct intel_dpll_hw_state *state);
 
 #endif /* __INTEL_DDI_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
index 3031e64ee518..8f23c4d51c33 100644
--- a/drivers/gpu/drm/i915/display/intel_display.c
+++ b/drivers/gpu/drm/i915/display/intel_display.c
@@ -41,7 +41,6 @@
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_rect.h>
-#include <drm/i915_drm.h>
 
 #include "display/intel_crt.h"
 #include "display/intel_ddi.h"
@@ -2720,9 +2719,10 @@ u32 intel_plane_fb_max_stride(struct drm_i915_private *dev_priv,
 
 	/*
 	 * We assume the primary plane for pipe A has
-	 * the highest stride limits of them all.
+	 * the highest stride limits of them all,
+	 * if in case pipe A is disabled, use the first pipe from pipe_mask.
 	 */
-	crtc = intel_get_crtc_for_pipe(dev_priv, PIPE_A);
+	crtc = intel_get_first_crtc(dev_priv);
 	if (!crtc)
 		return 0;
 
@@ -9542,7 +9542,7 @@ static void ilk_init_pch_refclk(struct drm_i915_private *dev_priv)
 	}
 
 	/* Check if any DPLLs are using the SSC source */
-	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
+	for (i = 0; i < dev_priv->dpll.num_shared_dpll; i++) {
 		u32 temp = intel_de_read(dev_priv, PCH_DPLL(i));
 
 		if (!(temp & DPLL_VCO_ENABLE))
@@ -10129,6 +10129,9 @@ static void bdw_set_pipemisc(const struct intel_crtc_state *crtc_state)
 					   BIT(PLANE_CURSOR))) == 0)
 		val |= PIPEMISC_HDR_MODE_PRECISION;
 
+	if (INTEL_GEN(dev_priv) >= 12)
+		val |= PIPEMISC_PIXEL_ROUNDING_TRUNC;
+
 	intel_de_write(dev_priv, PIPEMISC(crtc->pipe), val);
 }
 
@@ -14299,11 +14302,11 @@ verify_single_dpll_state(struct drm_i915_private *dev_priv,
 	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);
+				pipe_name(crtc->pipe), pll->active_mask);
 	else
 		I915_STATE_WARN(pll->active_mask & crtc_mask,
 				"pll active mismatch (didn't expect pipe %c in active mask 0x%02x)\n",
-				pipe_name(drm_crtc_index(&crtc->base)), pll->active_mask);
+				pipe_name(crtc->pipe), pll->active_mask);
 
 	I915_STATE_WARN(!(pll->state.crtc_mask & crtc_mask),
 			"pll enabled crtcs mismatch (expected 0x%x in 0x%02x)\n",
@@ -14332,10 +14335,10 @@ verify_shared_dpll_state(struct intel_crtc *crtc,
 
 		I915_STATE_WARN(pll->active_mask & crtc_mask,
 				"pll active mismatch (didn't expect pipe %c in active mask)\n",
-				pipe_name(drm_crtc_index(&crtc->base)));
+				pipe_name(crtc->pipe));
 		I915_STATE_WARN(pll->state.crtc_mask & crtc_mask,
 				"pll enabled crtcs mismatch (found %x in enabled mask)\n",
-				pipe_name(drm_crtc_index(&crtc->base)));
+				pipe_name(crtc->pipe));
 	}
 }
 
@@ -14359,8 +14362,10 @@ verify_disabled_dpll_state(struct drm_i915_private *dev_priv)
 {
 	int i;
 
-	for (i = 0; i < dev_priv->num_shared_dpll; i++)
-		verify_single_dpll_state(dev_priv, &dev_priv->shared_dplls[i], NULL, NULL);
+	for (i = 0; i < dev_priv->dpll.num_shared_dpll; i++)
+		verify_single_dpll_state(dev_priv,
+					 &dev_priv->dpll.shared_dplls[i],
+					 NULL, NULL);
 }
 
 static void
@@ -15318,7 +15323,6 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 	struct intel_crtc *crtc;
 	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
 	struct skl_ddb_entry entries[I915_MAX_PIPES] = {};
-	const u8 num_pipes = INTEL_NUM_PIPES(dev_priv);
 	u8 update_pipes = 0, modeset_pipes = 0;
 	int i;
 
@@ -15355,7 +15359,7 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 				continue;
 
 			if (skl_ddb_allocation_overlaps(&new_crtc_state->wm.skl.ddb,
-							entries, num_pipes, pipe))
+							entries, I915_MAX_PIPES, pipe))
 				continue;
 
 			entries[pipe] = new_crtc_state->wm.skl.ddb;
@@ -15393,7 +15397,7 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 			continue;
 
 		drm_WARN_ON(&dev_priv->drm, skl_ddb_allocation_overlaps(&new_crtc_state->wm.skl.ddb,
-									entries, num_pipes, pipe));
+									entries, I915_MAX_PIPES, pipe));
 
 		entries[pipe] = new_crtc_state->wm.skl.ddb;
 		modeset_pipes &= ~BIT(pipe);
@@ -15428,7 +15432,7 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 			continue;
 
 		drm_WARN_ON(&dev_priv->drm, skl_ddb_allocation_overlaps(&new_crtc_state->wm.skl.ddb,
-									entries, num_pipes, pipe));
+									entries, I915_MAX_PIPES, pipe));
 
 		entries[pipe] = new_crtc_state->wm.skl.ddb;
 		modeset_pipes &= ~BIT(pipe);
@@ -16320,7 +16324,6 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	struct intel_plane *plane;
 	const struct drm_plane_funcs *plane_funcs;
 	unsigned int supported_rotations;
-	unsigned int possible_crtcs;
 	const u32 *formats;
 	int num_formats;
 	int ret, zpos;
@@ -16401,18 +16404,16 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	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,
+					       0, plane_funcs,
 					       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,
+					       0, plane_funcs,
 					       formats, num_formats,
 					       i9xx_format_modifiers,
 					       DRM_PLANE_TYPE_PRIMARY,
@@ -16454,7 +16455,6 @@ static struct intel_plane *
 intel_cursor_plane_create(struct drm_i915_private *dev_priv,
 			  enum pipe pipe)
 {
-	unsigned int possible_crtcs;
 	struct intel_plane *cursor;
 	int ret, zpos;
 
@@ -16487,10 +16487,8 @@ intel_cursor_plane_create(struct drm_i915_private *dev_priv,
 	if (IS_I845G(dev_priv) || IS_I865G(dev_priv) || HAS_CUR_FBC(dev_priv))
 		cursor->cursor.size = ~0;
 
-	possible_crtcs = BIT(pipe);
-
 	ret = drm_universal_plane_init(&dev_priv->drm, &cursor->base,
-				       possible_crtcs, &intel_cursor_plane_funcs,
+				       0, &intel_cursor_plane_funcs,
 				       intel_cursor_formats,
 				       ARRAY_SIZE(intel_cursor_formats),
 				       cursor_format_modifiers,
@@ -16619,6 +16617,18 @@ static void intel_crtc_free(struct intel_crtc *crtc)
 	kfree(crtc);
 }
 
+static void intel_plane_possible_crtcs_init(struct drm_i915_private *dev_priv)
+{
+	struct intel_plane *plane;
+
+	for_each_intel_plane(&dev_priv->drm, plane) {
+		struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv,
+								  plane->pipe);
+
+		plane->base.possible_crtcs = drm_crtc_mask(&crtc->base);
+	}
+}
+
 static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
 	struct intel_plane *primary, *cursor;
@@ -16697,6 +16707,8 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 
 	intel_color_init(crtc);
 
+	intel_crtc_crc_init(crtc);
+
 	drm_WARN_ON(&dev_priv->drm, drm_crtc_index(&crtc->base) != crtc->pipe);
 
 	return 0;
@@ -17785,11 +17797,9 @@ static void plane_config_fini(struct intel_initial_plane_config *plane_config)
 		i915_vma_put(plane_config->vma);
 }
 
-int intel_modeset_init(struct drm_i915_private *i915)
+/* part #1: call before irq install */
+int intel_modeset_init_noirq(struct drm_i915_private *i915)
 {
-	struct drm_device *dev = &i915->drm;
-	enum pipe pipe;
-	struct intel_crtc *crtc;
 	int ret;
 
 	i915->modeset_wq = alloc_ordered_workqueue("i915_modeset", 0);
@@ -17814,6 +17824,17 @@ int intel_modeset_init(struct drm_i915_private *i915)
 
 	intel_fbc_init(i915);
 
+	return 0;
+}
+
+/* part #2: call after irq install */
+int intel_modeset_init(struct drm_i915_private *i915)
+{
+	struct drm_device *dev = &i915->drm;
+	enum pipe pipe;
+	struct intel_crtc *crtc;
+	int ret;
+
 	intel_init_pm(i915);
 
 	intel_panel_sanitize_ssc(i915);
@@ -17834,6 +17855,7 @@ int intel_modeset_init(struct drm_i915_private *i915)
 		}
 	}
 
+	intel_plane_possible_crtcs_init(i915);
 	intel_shared_dpll_init(dev);
 	intel_update_fdi_pll_freq(i915);
 
@@ -18311,7 +18333,6 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 	struct intel_connector *connector;
 	struct drm_connector_list_iter conn_iter;
 	u8 active_pipes = 0;
-	int i;
 
 	for_each_intel_crtc(dev, crtc) {
 		struct intel_crtc_state *crtc_state =
@@ -18340,33 +18361,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 
 	readout_plane_state(dev_priv);
 
-	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
-		struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i];
-
-		pll->on = pll->info->funcs->get_hw_state(dev_priv, pll,
-							&pll->state.hw_state);
-
-		if (IS_ELKHARTLAKE(dev_priv) && pll->on &&
-		    pll->info->id == DPLL_ID_EHL_DPLL4) {
-			pll->wakeref = intel_display_power_get(dev_priv,
-							       POWER_DOMAIN_DPLL_DC_OFF);
-		}
-
-		pll->state.crtc_mask = 0;
-		for_each_intel_crtc(dev, crtc) {
-			struct intel_crtc_state *crtc_state =
-				to_intel_crtc_state(crtc->base.state);
-
-			if (crtc_state->hw.active &&
-			    crtc_state->shared_dpll == pll)
-				pll->state.crtc_mask |= 1 << crtc->pipe;
-		}
-		pll->active_mask = pll->state.crtc_mask;
-
-		drm_dbg_kms(&dev_priv->drm,
-			    "%s hw state readout: crtc_mask 0x%08x, on %i\n",
-			    pll->info->name, pll->state.crtc_mask, pll->on);
-	}
+	intel_dpll_readout_hw_state(dev_priv);
 
 	for_each_intel_encoder(dev, encoder) {
 		pipe = 0;
@@ -18623,7 +18618,6 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
 	struct intel_encoder *encoder;
 	struct intel_crtc *crtc;
 	intel_wakeref_t wakeref;
-	int i;
 
 	wakeref = intel_display_power_get(dev_priv, POWER_DOMAIN_INIT);
 
@@ -18676,19 +18670,7 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
 
 	intel_modeset_update_connector_atomic_state(dev);
 
-	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
-		struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i];
-
-		if (!pll->on || pll->active_mask)
-			continue;
-
-		drm_dbg_kms(&dev_priv->drm,
-			    "%s enabled but not in use, disabling\n",
-			    pll->info->name);
-
-		pll->info->funcs->disable(dev_priv, pll);
-		pll->on = false;
-	}
+	intel_dpll_sanitize_state(dev_priv);
 
 	if (IS_G4X(dev_priv)) {
 		g4x_wm_get_hw_state(dev_priv);
@@ -18820,6 +18802,15 @@ void intel_modeset_driver_remove_noirq(struct drm_i915_private *i915)
 
 #if IS_ENABLED(CONFIG_DRM_I915_CAPTURE_ERROR)
 
+static bool
+has_transcoder(struct drm_i915_private *dev_priv, enum transcoder cpu_transcoder)
+{
+	if (cpu_transcoder == TRANSCODER_EDP)
+		return HAS_TRANSCODER_EDP(dev_priv);
+	else
+		return INTEL_INFO(dev_priv)->pipe_mask & BIT(cpu_transcoder);
+}
+
 struct intel_display_error_state {
 
 	u32 power_well_driver;
@@ -18928,7 +18919,7 @@ intel_display_capture_error_state(struct drm_i915_private *dev_priv)
 	for (i = 0; i < ARRAY_SIZE(error->transcoder); i++) {
 		enum transcoder cpu_transcoder = transcoders[i];
 
-		if (!INTEL_INFO(dev_priv)->trans_offsets[cpu_transcoder])
+		if (!has_transcoder(dev_priv, cpu_transcoder))
 			continue;
 
 		error->transcoder[i].available = true;
diff --git a/drivers/gpu/drm/i915/display/intel_display.h b/drivers/gpu/drm/i915/display/intel_display.h
index f92efbbec838..adb1225a3480 100644
--- a/drivers/gpu/drm/i915/display/intel_display.h
+++ b/drivers/gpu/drm/i915/display/intel_display.h
@@ -26,7 +26,6 @@
 #define _INTEL_DISPLAY_H_
 
 #include <drm/drm_util.h>
-#include <drm/i915_drm.h>
 
 enum link_m_n_set;
 struct dpll;
@@ -40,6 +39,7 @@ struct drm_framebuffer;
 struct drm_i915_error_state_buf;
 struct drm_i915_gem_object;
 struct drm_i915_private;
+struct drm_mode_fb_cmd2;
 struct drm_modeset_acquire_ctx;
 struct drm_plane;
 struct drm_plane_state;
@@ -47,6 +47,7 @@ struct i915_ggtt_view;
 struct intel_atomic_state;
 struct intel_crtc;
 struct intel_crtc_state;
+struct intel_crtc_state;
 struct intel_digital_port;
 struct intel_dp;
 struct intel_encoder;
@@ -55,7 +56,6 @@ struct intel_plane;
 struct intel_plane_state;
 struct intel_remapped_info;
 struct intel_rotation_info;
-struct intel_crtc_state;
 
 enum i915_gpio {
 	GPIOA,
@@ -313,10 +313,11 @@ enum phy_fia {
 };
 
 #define for_each_pipe(__dev_priv, __p) \
-	for ((__p) = 0; (__p) < INTEL_NUM_PIPES(__dev_priv); (__p)++)
+	for ((__p) = 0; (__p) < I915_MAX_PIPES; (__p)++) \
+		for_each_if(INTEL_INFO(__dev_priv)->pipe_mask & BIT(__p))
 
 #define for_each_pipe_masked(__dev_priv, __p, __mask) \
-	for ((__p) = 0; (__p) < INTEL_NUM_PIPES(__dev_priv); (__p)++) \
+	for_each_pipe(__dev_priv, __p) \
 		for_each_if((__mask) & BIT(__p))
 
 #define for_each_cpu_transcoder_masked(__dev_priv, __t, __mask) \
@@ -614,6 +615,7 @@ intel_format_info_is_yuv_semiplanar(const struct drm_format_info *info,
 
 /* modesetting */
 void intel_modeset_init_hw(struct drm_i915_private *i915);
+int intel_modeset_init_noirq(struct drm_i915_private *i915);
 int intel_modeset_init(struct drm_i915_private *i915);
 void intel_modeset_driver_remove(struct drm_i915_private *i915);
 void intel_modeset_driver_remove_noirq(struct drm_i915_private *i915);
diff --git a/drivers/gpu/drm/i915/display/intel_display_debugfs.c b/drivers/gpu/drm/i915/display/intel_display_debugfs.c
index 46954cc7b6c0..1e6eb7f2f72d 100644
--- a/drivers/gpu/drm/i915/display/intel_display_debugfs.c
+++ b/drivers/gpu/drm/i915/display/intel_display_debugfs.c
@@ -920,8 +920,13 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 	int i;
 
 	drm_modeset_lock_all(dev);
-	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
-		struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i];
+
+	seq_printf(m, "PLL refclks: non-SSC: %d kHz, SSC: %d kHz\n",
+		   dev_priv->dpll.ref_clks.nssc,
+		   dev_priv->dpll.ref_clks.ssc);
+
+	for (i = 0; i < dev_priv->dpll.num_shared_dpll; i++) {
+		struct intel_shared_dpll *pll = &dev_priv->dpll.shared_dplls[i];
 
 		seq_printf(m, "DPLL%i: %s, id: %i\n", i, pll->info->name,
 			   pll->info->id);
diff --git a/drivers/gpu/drm/i915/display/intel_display_power.c b/drivers/gpu/drm/i915/display/intel_display_power.c
index 6e25a1317161..246e406bb385 100644
--- a/drivers/gpu/drm/i915/display/intel_display_power.c
+++ b/drivers/gpu/drm/i915/display/intel_display_power.c
@@ -939,11 +939,17 @@ unlock:
 
 static void assert_can_enable_dc5(struct drm_i915_private *dev_priv)
 {
-	bool pg2_enabled = intel_display_power_well_is_enabled(dev_priv,
-					SKL_DISP_PW_2);
+	enum i915_power_well_id high_pg;
 
-	drm_WARN_ONCE(&dev_priv->drm, pg2_enabled,
-		      "PG2 not disabled to enable DC5.\n");
+	/* Power wells at this level and above must be disabled for DC5 entry */
+	if (INTEL_GEN(dev_priv) >= 12)
+		high_pg = TGL_DISP_PW_3;
+	else
+		high_pg = SKL_DISP_PW_2;
+
+	drm_WARN_ONCE(&dev_priv->drm,
+		      intel_display_power_well_is_enabled(dev_priv, high_pg),
+		      "Power wells above platform's DC5 limit still enabled.\n");
 
 	drm_WARN_ONCE(&dev_priv->drm,
 		      (intel_de_read(dev_priv, DC_STATE_EN) &
@@ -2740,7 +2746,7 @@ void intel_display_power_put(struct drm_i915_private *dev_priv,
 	BIT_ULL(POWER_DOMAIN_INIT))
 
 #define TGL_DISPLAY_DC_OFF_POWER_DOMAINS (		\
-	TGL_PW_2_POWER_DOMAINS |			\
+	TGL_PW_3_POWER_DOMAINS |			\
 	BIT_ULL(POWER_DOMAIN_MODESET) |			\
 	BIT_ULL(POWER_DOMAIN_AUX_A) |			\
 	BIT_ULL(POWER_DOMAIN_AUX_B) |			\
@@ -3936,7 +3942,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 		.name = "power well 3",
 		.domains = TGL_PW_3_POWER_DOMAINS,
 		.ops = &hsw_power_well_ops,
-		.id = DISP_PW_ID_NONE,
+		.id = TGL_DISP_PW_3,
 		{
 			.hsw.regs = &hsw_power_well_regs,
 			.hsw.idx = ICL_PW_CTL_IDX_PW_3,
diff --git a/drivers/gpu/drm/i915/display/intel_display_power.h b/drivers/gpu/drm/i915/display/intel_display_power.h
index 601e000ffd0d..da64a5edae7a 100644
--- a/drivers/gpu/drm/i915/display/intel_display_power.h
+++ b/drivers/gpu/drm/i915/display/intel_display_power.h
@@ -100,6 +100,7 @@ enum i915_power_well_id {
 	SKL_DISP_PW_MISC_IO,
 	SKL_DISP_PW_1,
 	SKL_DISP_PW_2,
+	TGL_DISP_PW_3,
 	SKL_DISP_DC_OFF,
 };
 
diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h b/drivers/gpu/drm/i915/display/intel_display_types.h
index 0d8a64305464..5e00e611f077 100644
--- a/drivers/gpu/drm/i915/display/intel_display_types.h
+++ b/drivers/gpu/drm/i915/display/intel_display_types.h
@@ -39,7 +39,6 @@
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_rect.h>
 #include <drm/drm_vblank.h>
-#include <drm/i915_drm.h>
 #include <drm/i915_mei_hdcp_interface.h>
 #include <media/cec-notifier.h>
 
@@ -642,6 +641,14 @@ struct intel_crtc_scaler_state {
 /* Flag to use the scanline counter instead of the pixel counter */
 #define I915_MODE_FLAG_USE_SCANLINE_COUNTER (1<<2)
 
+struct intel_wm_level {
+	bool enable;
+	u32 pri_val;
+	u32 spr_val;
+	u32 cur_val;
+	u32 fbc_val;
+};
+
 struct intel_pipe_wm {
 	struct intel_wm_level wm[5];
 	bool fbc_wm_enabled;
@@ -650,6 +657,14 @@ struct intel_pipe_wm {
 	bool sprites_scaled;
 };
 
+struct skl_wm_level {
+	u16 min_ddb_alloc;
+	u16 plane_res_b;
+	u8 plane_res_l;
+	bool plane_en;
+	bool ignore_lines;
+};
+
 struct skl_plane_wm {
 	struct skl_wm_level wm[8];
 	struct skl_wm_level uv_wm[8];
@@ -1046,6 +1061,32 @@ struct intel_crtc_state {
 	enum transcoder mst_master_transcoder;
 };
 
+enum intel_pipe_crc_source {
+	INTEL_PIPE_CRC_SOURCE_NONE,
+	INTEL_PIPE_CRC_SOURCE_PLANE1,
+	INTEL_PIPE_CRC_SOURCE_PLANE2,
+	INTEL_PIPE_CRC_SOURCE_PLANE3,
+	INTEL_PIPE_CRC_SOURCE_PLANE4,
+	INTEL_PIPE_CRC_SOURCE_PLANE5,
+	INTEL_PIPE_CRC_SOURCE_PLANE6,
+	INTEL_PIPE_CRC_SOURCE_PLANE7,
+	INTEL_PIPE_CRC_SOURCE_PIPE,
+	/* TV/DP on pre-gen5/vlv can't use the pipe source. */
+	INTEL_PIPE_CRC_SOURCE_TV,
+	INTEL_PIPE_CRC_SOURCE_DP_B,
+	INTEL_PIPE_CRC_SOURCE_DP_C,
+	INTEL_PIPE_CRC_SOURCE_DP_D,
+	INTEL_PIPE_CRC_SOURCE_AUTO,
+	INTEL_PIPE_CRC_SOURCE_MAX,
+};
+
+#define INTEL_PIPE_CRC_ENTRIES_NR	128
+struct intel_pipe_crc {
+	spinlock_t lock;
+	int skipped;
+	enum intel_pipe_crc_source source;
+};
+
 struct intel_crtc {
 	struct drm_crtc base;
 	enum pipe pipe;
@@ -1089,6 +1130,10 @@ struct intel_crtc {
 
 	/* per pipe DSB related info */
 	struct intel_dsb dsb;
+
+#ifdef CONFIG_DEBUG_FS
+	struct intel_pipe_crc pipe_crc;
+#endif
 };
 
 struct intel_plane {
@@ -1235,6 +1280,7 @@ struct intel_dp {
 	int max_link_rate;
 	/* sink or branch descriptor */
 	struct drm_dp_desc desc;
+	u32 edid_quirks;
 	struct drm_dp_aux aux;
 	u32 aux_busy_last_status;
 	u8 train_set[4];
@@ -1407,8 +1453,17 @@ vlv_pipe_to_channel(enum pipe pipe)
 }
 
 static inline struct intel_crtc *
+intel_get_first_crtc(struct drm_i915_private *dev_priv)
+{
+	return to_intel_crtc(drm_crtc_from_index(&dev_priv->drm, 0));
+}
+
+static inline struct intel_crtc *
 intel_get_crtc_for_pipe(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
+	/* pipe_to_crtc_mapping may have hole on any of 3 display pipe system */
+	drm_WARN_ON(&dev_priv->drm,
+		    !(INTEL_INFO(dev_priv)->pipe_mask & BIT(pipe)));
 	return dev_priv->pipe_to_crtc_mapping[pipe];
 }
 
@@ -1598,11 +1653,15 @@ intel_crtc_has_dp_encoder(const struct intel_crtc_state *crtc_state)
 		 (1 << INTEL_OUTPUT_DP_MST) |
 		 (1 << INTEL_OUTPUT_EDP));
 }
+
 static inline void
 intel_wait_for_vblank(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
-	drm_wait_one_vblank(&dev_priv->drm, pipe);
+	struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
+
+	drm_crtc_wait_one_vblank(&crtc->base);
 }
+
 static inline void
 intel_wait_for_vblank_if_active(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
index 0a5a9197f8f5..0a417cd2af2b 100644
--- a/drivers/gpu/drm/i915/display/intel_dp.c
+++ b/drivers/gpu/drm/i915/display/intel_dp.c
@@ -40,7 +40,6 @@
 #include <drm/drm_edid.h>
 #include <drm/drm_hdcp.h>
 #include <drm/drm_probe_helper.h>
-#include <drm/i915_drm.h>
 
 #include "i915_debugfs.h"
 #include "i915_drv.h"
@@ -2399,7 +2398,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
 	struct intel_digital_connector_state *intel_conn_state =
 		to_intel_digital_connector_state(conn_state);
-	bool constant_n = drm_dp_has_quirk(&intel_dp->desc,
+	bool constant_n = drm_dp_has_quirk(&intel_dp->desc, 0,
 					   DP_DPCD_QUIRK_CONSTANT_N);
 	int ret = 0, output_bpp;
 
@@ -4515,7 +4514,8 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	 * it don't care about read it here and in intel_edp_init_dpcd().
 	 */
 	if (!intel_dp_is_edp(intel_dp) &&
-	    !drm_dp_has_quirk(&intel_dp->desc, DP_DPCD_QUIRK_NO_SINK_COUNT)) {
+	    !drm_dp_has_quirk(&intel_dp->desc, 0,
+			      DP_DPCD_QUIRK_NO_SINK_COUNT)) {
 		u8 count;
 		ssize_t r;
 
@@ -5682,6 +5682,7 @@ intel_dp_set_edid(struct intel_dp *intel_dp)
 
 	intel_dp->has_audio = drm_detect_monitor_audio(edid);
 	drm_dp_cec_set_edid(&intel_dp->aux, edid);
+	intel_dp->edid_quirks = drm_dp_get_edid_quirks(edid);
 }
 
 static void
@@ -5694,6 +5695,7 @@ intel_dp_unset_edid(struct intel_dp *intel_dp)
 	intel_connector->detect_edid = NULL;
 
 	intel_dp->has_audio = false;
+	intel_dp->edid_quirks = 0;
 }
 
 static int
@@ -6449,6 +6451,7 @@ static
 int intel_dp_hdcp2_config_stream_type(struct intel_digital_port *intel_dig_port,
 				      bool is_repeater, u8 content_type)
 {
+	int ret;
 	struct hdcp2_dp_errata_stream_type stream_type_msg;
 
 	if (is_repeater)
@@ -6464,8 +6467,11 @@ int intel_dp_hdcp2_config_stream_type(struct intel_digital_port *intel_dig_port,
 	stream_type_msg.msg_id = HDCP_2_2_ERRATA_DP_STREAM_TYPE;
 	stream_type_msg.stream_type = content_type;
 
-	return intel_dp_hdcp2_write_msg(intel_dig_port, &stream_type_msg,
+	ret =  intel_dp_hdcp2_write_msg(intel_dig_port, &stream_type_msg,
 					sizeof(stream_type_msg));
+
+	return ret < 0 ? ret : 0;
+
 }
 
 static
@@ -7562,8 +7568,8 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	edid = drm_get_edid(connector, &intel_dp->aux.ddc);
 	if (edid) {
 		if (drm_add_edid_modes(connector, edid)) {
-			drm_connector_update_edid_property(connector,
-								edid);
+			drm_connector_update_edid_property(connector, edid);
+			intel_dp->edid_quirks = drm_dp_get_edid_quirks(edid);
 		} else {
 			kfree(edid);
 			edid = ERR_PTR(-EINVAL);
@@ -7609,9 +7615,8 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	intel_panel_setup_backlight(connector, pipe);
 
 	if (fixed_mode) {
-		/* We do not know the orientation, but their might be a quirk */
 		drm_connector_set_panel_orientation_with_quirk(connector,
-				DRM_MODE_PANEL_ORIENTATION_UNKNOWN,
+				dev_priv->vbt.orientation,
 				fixed_mode->hdisplay, fixed_mode->vdisplay);
 	}
 
diff --git a/drivers/gpu/drm/i915/display/intel_dp.h b/drivers/gpu/drm/i915/display/intel_dp.h
index 3da166054788..0c7be8ed1423 100644
--- a/drivers/gpu/drm/i915/display/intel_dp.h
+++ b/drivers/gpu/drm/i915/display/intel_dp.h
@@ -8,8 +8,6 @@
 
 #include <linux/types.h>
 
-#include <drm/i915_drm.h>
-
 #include "i915_reg.h"
 
 enum pipe;
diff --git a/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c b/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
index 48276237b362..3e706bb850a8 100644
--- a/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
+++ b/drivers/gpu/drm/i915/display/intel_dp_aux_backlight.c
@@ -328,15 +328,31 @@ intel_dp_aux_display_control_capable(struct intel_connector *connector)
 int intel_dp_aux_init_backlight_funcs(struct intel_connector *intel_connector)
 {
 	struct intel_panel *panel = &intel_connector->panel;
-	struct drm_i915_private *dev_priv = to_i915(intel_connector->base.dev);
+	struct intel_dp *intel_dp = enc_to_intel_dp(intel_connector->encoder);
+	struct drm_device *dev = intel_connector->base.dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
 
 	if (i915_modparams.enable_dpcd_backlight == 0 ||
-	    (i915_modparams.enable_dpcd_backlight == -1 &&
-	    dev_priv->vbt.backlight.type != INTEL_BACKLIGHT_VESA_EDP_AUX_INTERFACE))
+	    !intel_dp_aux_display_control_capable(intel_connector))
 		return -ENODEV;
 
-	if (!intel_dp_aux_display_control_capable(intel_connector))
+	/*
+	 * There are a lot of machines that don't advertise the backlight
+	 * control interface to use properly in their VBIOS, :\
+	 */
+	if (dev_priv->vbt.backlight.type !=
+	    INTEL_BACKLIGHT_VESA_EDP_AUX_INTERFACE &&
+	    !drm_dp_has_quirk(&intel_dp->desc, intel_dp->edid_quirks,
+			      DP_QUIRK_FORCE_DPCD_BACKLIGHT)) {
+		DRM_DEV_INFO(dev->dev,
+			     "Panel advertises DPCD backlight support, but "
+			     "VBT disagrees. If your backlight controls "
+			     "don't work try booting with "
+			     "i915.enable_dpcd_backlight=1. If your machine "
+			     "needs this, please file a _new_ bug report on "
+			     "drm/i915, see " FDO_BUG_URL " for details.\n");
 		return -ENODEV;
+	}
 
 	panel->backlight.setup = intel_dp_aux_setup_backlight;
 	panel->backlight.enable = intel_dp_aux_enable_backlight;
diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c
index b15404a3b1ca..e08caca658c6 100644
--- a/drivers/gpu/drm/i915/display/intel_dp_mst.c
+++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c
@@ -50,7 +50,7 @@ static int intel_dp_mst_compute_link_config(struct intel_encoder *encoder,
 	const struct drm_display_mode *adjusted_mode =
 		&crtc_state->hw.adjusted_mode;
 	void *port = connector->port;
-	bool constant_n = drm_dp_has_quirk(&intel_dp->desc,
+	bool constant_n = drm_dp_has_quirk(&intel_dp->desc, 0,
 					   DP_DPCD_QUIRK_CONSTANT_N);
 	int bpp, slots = -EINVAL;
 
@@ -548,12 +548,41 @@ static int intel_dp_mst_get_ddc_modes(struct drm_connector *connector)
 	return ret;
 }
 
+static int
+intel_dp_mst_connector_late_register(struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	int ret;
+
+	ret = drm_dp_mst_connector_late_register(connector,
+						 intel_connector->port);
+	if (ret < 0)
+		return ret;
+
+	ret = intel_connector_register(connector);
+	if (ret < 0)
+		drm_dp_mst_connector_early_unregister(connector,
+						      intel_connector->port);
+
+	return ret;
+}
+
+static void
+intel_dp_mst_connector_early_unregister(struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+
+	intel_connector_unregister(connector);
+	drm_dp_mst_connector_early_unregister(connector,
+					      intel_connector->port);
+}
+
 static const struct drm_connector_funcs intel_dp_mst_connector_funcs = {
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.atomic_get_property = intel_digital_connector_atomic_get_property,
 	.atomic_set_property = intel_digital_connector_atomic_set_property,
-	.late_register = intel_connector_register,
-	.early_unregister = intel_connector_unregister,
+	.late_register = intel_dp_mst_connector_late_register,
+	.early_unregister = intel_dp_mst_connector_early_unregister,
 	.destroy = intel_connector_destroy,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
 	.atomic_duplicate_state = intel_digital_connector_duplicate_state,
diff --git a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
index e5bfe5245276..2d47f1f756a2 100644
--- a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
@@ -45,6 +45,22 @@
  * commit phase.
  */
 
+struct intel_dpll_mgr {
+	const struct dpll_info *dpll_info;
+
+	bool (*get_dplls)(struct intel_atomic_state *state,
+			  struct intel_crtc *crtc,
+			  struct intel_encoder *encoder);
+	void (*put_dplls)(struct intel_atomic_state *state,
+			  struct intel_crtc *crtc);
+	void (*update_active_dpll)(struct intel_atomic_state *state,
+				   struct intel_crtc *crtc,
+				   struct intel_encoder *encoder);
+	void (*update_ref_clks)(struct drm_i915_private *i915);
+	void (*dump_hw_state)(struct drm_i915_private *dev_priv,
+			      const struct intel_dpll_hw_state *hw_state);
+};
+
 static void
 intel_atomic_duplicate_dpll_state(struct drm_i915_private *dev_priv,
 				  struct intel_shared_dpll_state *shared_dpll)
@@ -52,8 +68,8 @@ intel_atomic_duplicate_dpll_state(struct drm_i915_private *dev_priv,
 	enum intel_dpll_id i;
 
 	/* Copy shared dpll state */
-	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
-		struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i];
+	for (i = 0; i < dev_priv->dpll.num_shared_dpll; i++) {
+		struct intel_shared_dpll *pll = &dev_priv->dpll.shared_dplls[i];
 
 		shared_dpll[i] = pll->state;
 	}
@@ -88,7 +104,7 @@ struct intel_shared_dpll *
 intel_get_shared_dpll_by_id(struct drm_i915_private *dev_priv,
 			    enum intel_dpll_id id)
 {
-	return &dev_priv->shared_dplls[id];
+	return &dev_priv->dpll.shared_dplls[id];
 }
 
 /**
@@ -103,11 +119,14 @@ enum intel_dpll_id
 intel_get_shared_dpll_id(struct drm_i915_private *dev_priv,
 			 struct intel_shared_dpll *pll)
 {
-	if (drm_WARN_ON(&dev_priv->drm, pll < dev_priv->shared_dplls ||
-			pll > &dev_priv->shared_dplls[dev_priv->num_shared_dpll]))
+	long pll_idx = pll - dev_priv->dpll.shared_dplls;
+
+	if (drm_WARN_ON(&dev_priv->drm,
+			pll_idx < 0 ||
+			pll_idx >= dev_priv->dpll.num_shared_dpll))
 		return -1;
 
-	return (enum intel_dpll_id) (pll - dev_priv->shared_dplls);
+	return pll_idx;
 }
 
 /* For ILK+ */
@@ -144,7 +163,7 @@ void intel_prepare_shared_dpll(const struct intel_crtc_state *crtc_state)
 	if (drm_WARN_ON(&dev_priv->drm, pll == NULL))
 		return;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 	drm_WARN_ON(&dev_priv->drm, !pll->state.crtc_mask);
 	if (!pll->active_mask) {
 		drm_dbg(&dev_priv->drm, "setting up %s\n", pll->info->name);
@@ -153,7 +172,7 @@ void intel_prepare_shared_dpll(const struct intel_crtc_state *crtc_state)
 
 		pll->info->funcs->prepare(dev_priv, pll);
 	}
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 /**
@@ -173,7 +192,7 @@ void intel_enable_shared_dpll(const struct intel_crtc_state *crtc_state)
 	if (drm_WARN_ON(&dev_priv->drm, pll == NULL))
 		return;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 	old_mask = pll->active_mask;
 
 	if (drm_WARN_ON(&dev_priv->drm, !(pll->state.crtc_mask & crtc_mask)) ||
@@ -199,7 +218,7 @@ void intel_enable_shared_dpll(const struct intel_crtc_state *crtc_state)
 	pll->on = true;
 
 out:
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 /**
@@ -222,7 +241,7 @@ void intel_disable_shared_dpll(const struct intel_crtc_state *crtc_state)
 	if (pll == NULL)
 		return;
 
-	mutex_lock(&dev_priv->dpll_lock);
+	mutex_lock(&dev_priv->dpll.lock);
 	if (drm_WARN_ON(&dev_priv->drm, !(pll->active_mask & crtc_mask)))
 		goto out;
 
@@ -243,7 +262,7 @@ void intel_disable_shared_dpll(const struct intel_crtc_state *crtc_state)
 	pll->on = false;
 
 out:
-	mutex_unlock(&dev_priv->dpll_lock);
+	mutex_unlock(&dev_priv->dpll.lock);
 }
 
 static struct intel_shared_dpll *
@@ -262,7 +281,7 @@ intel_find_shared_dpll(struct intel_atomic_state *state,
 	drm_WARN_ON(&dev_priv->drm, dpll_mask & ~(BIT(I915_NUM_PLLS) - 1));
 
 	for_each_set_bit(i, &dpll_mask, I915_NUM_PLLS) {
-		pll = &dev_priv->shared_dplls[i];
+		pll = &dev_priv->dpll.shared_dplls[i];
 
 		/* Only want to check enabled timings first */
 		if (shared_dpll[i].crtc_mask == 0) {
@@ -362,9 +381,9 @@ void intel_shared_dpll_swap_state(struct intel_atomic_state *state)
 	if (!state->dpll_set)
 		return;
 
-	for (i = 0; i < dev_priv->num_shared_dpll; i++) {
+	for (i = 0; i < dev_priv->dpll.num_shared_dpll; i++) {
 		struct intel_shared_dpll *pll =
-			&dev_priv->shared_dplls[i];
+			&dev_priv->dpll.shared_dplls[i];
 
 		swap(pll->state, shared_dpll[i]);
 	}
@@ -462,7 +481,7 @@ static bool ibx_get_dpll(struct intel_atomic_state *state,
 	if (HAS_PCH_IBX(dev_priv)) {
 		/* Ironlake PCH has a fixed PLL->PCH pipe mapping. */
 		i = (enum intel_dpll_id) crtc->pipe;
-		pll = &dev_priv->shared_dplls[i];
+		pll = &dev_priv->dpll.shared_dplls[i];
 
 		drm_dbg_kms(&dev_priv->drm,
 			    "[CRTC:%d:%s] using pre-allocated %s\n",
@@ -506,6 +525,19 @@ static const struct intel_shared_dpll_funcs ibx_pch_dpll_funcs = {
 	.get_hw_state = ibx_pch_dpll_get_hw_state,
 };
 
+static const struct dpll_info pch_plls[] = {
+	{ "PCH DPLL A", &ibx_pch_dpll_funcs, DPLL_ID_PCH_PLL_A, 0 },
+	{ "PCH DPLL B", &ibx_pch_dpll_funcs, DPLL_ID_PCH_PLL_B, 0 },
+	{ },
+};
+
+static const struct intel_dpll_mgr pch_pll_mgr = {
+	.dpll_info = pch_plls,
+	.get_dplls = ibx_get_dpll,
+	.put_dplls = intel_put_dpll,
+	.dump_hw_state = ibx_dump_hw_state,
+};
+
 static void hsw_ddi_wrpll_enable(struct drm_i915_private *dev_priv,
 			       struct intel_shared_dpll *pll)
 {
@@ -818,8 +850,8 @@ hsw_ddi_calculate_wrpll(int clock /* in Hz */,
 }
 
 static struct intel_shared_dpll *
-hsw_ddi_hdmi_get_dpll(struct intel_atomic_state *state,
-		      struct intel_crtc *crtc)
+hsw_ddi_wrpll_get_dpll(struct intel_atomic_state *state,
+		       struct intel_crtc *crtc)
 {
 	struct intel_crtc_state *crtc_state =
 		intel_atomic_get_new_crtc_state(state, crtc);
@@ -846,8 +878,47 @@ hsw_ddi_hdmi_get_dpll(struct intel_atomic_state *state,
 	return pll;
 }
 
+static int hsw_ddi_wrpll_get_freq(struct drm_i915_private *dev_priv,
+				  const struct intel_shared_dpll *pll)
+{
+	int refclk;
+	int n, p, r;
+	u32 wrpll = pll->state.hw_state.wrpll;
+
+	switch (wrpll & WRPLL_REF_MASK) {
+	case WRPLL_REF_SPECIAL_HSW:
+		/* Muxed-SSC for BDW, non-SSC for non-ULT HSW. */
+		if (IS_HASWELL(dev_priv) && !IS_HSW_ULT(dev_priv)) {
+			refclk = dev_priv->dpll.ref_clks.nssc;
+			break;
+		}
+		/* fall through */
+	case WRPLL_REF_PCH_SSC:
+		/*
+		 * We could calculate spread here, but our checking
+		 * code only cares about 5% accuracy, and spread is a max of
+		 * 0.5% downspread.
+		 */
+		refclk = dev_priv->dpll.ref_clks.ssc;
+		break;
+	case WRPLL_REF_LCPLL:
+		refclk = 2700000;
+		break;
+	default:
+		MISSING_CASE(wrpll);
+		return 0;
+	}
+
+	r = wrpll & WRPLL_DIVIDER_REF_MASK;
+	p = (wrpll & WRPLL_DIVIDER_POST_MASK) >> WRPLL_DIVIDER_POST_SHIFT;
+	n = (wrpll & WRPLL_DIVIDER_FB_MASK) >> WRPLL_DIVIDER_FB_SHIFT;
+
+	/* Convert to KHz, p & r have a fixed point portion */
+	return (refclk * n / 10) / (p * r) * 2;
+}
+
 static struct intel_shared_dpll *
-hsw_ddi_dp_get_dpll(struct intel_crtc_state *crtc_state)
+hsw_ddi_lcpll_get_dpll(struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	struct intel_shared_dpll *pll;
@@ -878,6 +949,69 @@ hsw_ddi_dp_get_dpll(struct intel_crtc_state *crtc_state)
 	return pll;
 }
 
+static int hsw_ddi_lcpll_get_freq(struct drm_i915_private *i915,
+				  const struct intel_shared_dpll *pll)
+{
+	int link_clock = 0;
+
+	switch (pll->info->id) {
+	case DPLL_ID_LCPLL_810:
+		link_clock = 81000;
+		break;
+	case DPLL_ID_LCPLL_1350:
+		link_clock = 135000;
+		break;
+	case DPLL_ID_LCPLL_2700:
+		link_clock = 270000;
+		break;
+	default:
+		drm_WARN(&i915->drm, 1, "bad port clock sel\n");
+		break;
+	}
+
+	return link_clock * 2;
+}
+
+static struct intel_shared_dpll *
+hsw_ddi_spll_get_dpll(struct intel_atomic_state *state,
+		      struct intel_crtc *crtc)
+{
+	struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+
+	if (WARN_ON(crtc_state->port_clock / 2 != 135000))
+		return NULL;
+
+	crtc_state->dpll_hw_state.spll = SPLL_PLL_ENABLE | SPLL_FREQ_1350MHz |
+					 SPLL_REF_MUXED_SSC;
+
+	return intel_find_shared_dpll(state, crtc, &crtc_state->dpll_hw_state,
+				      BIT(DPLL_ID_SPLL));
+}
+
+static int hsw_ddi_spll_get_freq(struct drm_i915_private *i915,
+				 const struct intel_shared_dpll *pll)
+{
+	int link_clock = 0;
+
+	switch (pll->state.hw_state.spll & SPLL_FREQ_MASK) {
+	case SPLL_FREQ_810MHz:
+		link_clock = 81000;
+		break;
+	case SPLL_FREQ_1350MHz:
+		link_clock = 135000;
+		break;
+	case SPLL_FREQ_2700MHz:
+		link_clock = 270000;
+		break;
+	default:
+		drm_WARN(&i915->drm, 1, "bad spll freq\n");
+		break;
+	}
+
+	return link_clock * 2;
+}
+
 static bool hsw_get_dpll(struct intel_atomic_state *state,
 			 struct intel_crtc *crtc,
 			 struct intel_encoder *encoder)
@@ -889,23 +1023,14 @@ static bool hsw_get_dpll(struct intel_atomic_state *state,
 	memset(&crtc_state->dpll_hw_state, 0,
 	       sizeof(crtc_state->dpll_hw_state));
 
-	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI)) {
-		pll = hsw_ddi_hdmi_get_dpll(state, crtc);
-	} else if (intel_crtc_has_dp_encoder(crtc_state)) {
-		pll = hsw_ddi_dp_get_dpll(crtc_state);
-	} else if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_ANALOG)) {
-		if (WARN_ON(crtc_state->port_clock / 2 != 135000))
-			return false;
-
-		crtc_state->dpll_hw_state.spll =
-			SPLL_PLL_ENABLE | SPLL_FREQ_1350MHz | SPLL_REF_MUXED_SSC;
-
-		pll = intel_find_shared_dpll(state, crtc,
-					     &crtc_state->dpll_hw_state,
-					     BIT(DPLL_ID_SPLL));
-	} else {
+	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI))
+		pll = hsw_ddi_wrpll_get_dpll(state, crtc);
+	else if (intel_crtc_has_dp_encoder(crtc_state))
+		pll = hsw_ddi_lcpll_get_dpll(crtc_state);
+	else if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_ANALOG))
+		pll = hsw_ddi_spll_get_dpll(state, crtc);
+	else
 		return false;
-	}
 
 	if (!pll)
 		return false;
@@ -918,6 +1043,16 @@ static bool hsw_get_dpll(struct intel_atomic_state *state,
 	return true;
 }
 
+static void hsw_update_dpll_ref_clks(struct drm_i915_private *i915)
+{
+	i915->dpll.ref_clks.ssc = 135000;
+	/* Non-SSC is only used on non-ULT HSW. */
+	if (intel_de_read(i915, FUSE_STRAP3) & HSW_REF_CLK_SELECT)
+		i915->dpll.ref_clks.nssc = 24000;
+	else
+		i915->dpll.ref_clks.nssc = 135000;
+}
+
 static void hsw_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state)
 {
@@ -929,12 +1064,14 @@ static const struct intel_shared_dpll_funcs hsw_ddi_wrpll_funcs = {
 	.enable = hsw_ddi_wrpll_enable,
 	.disable = hsw_ddi_wrpll_disable,
 	.get_hw_state = hsw_ddi_wrpll_get_hw_state,
+	.get_freq = hsw_ddi_wrpll_get_freq,
 };
 
 static const struct intel_shared_dpll_funcs hsw_ddi_spll_funcs = {
 	.enable = hsw_ddi_spll_enable,
 	.disable = hsw_ddi_spll_disable,
 	.get_hw_state = hsw_ddi_spll_get_hw_state,
+	.get_freq = hsw_ddi_spll_get_freq,
 };
 
 static void hsw_ddi_lcpll_enable(struct drm_i915_private *dev_priv,
@@ -958,6 +1095,25 @@ static const struct intel_shared_dpll_funcs hsw_ddi_lcpll_funcs = {
 	.enable = hsw_ddi_lcpll_enable,
 	.disable = hsw_ddi_lcpll_disable,
 	.get_hw_state = hsw_ddi_lcpll_get_hw_state,
+	.get_freq = hsw_ddi_lcpll_get_freq,
+};
+
+static const struct dpll_info hsw_plls[] = {
+	{ "WRPLL 1",    &hsw_ddi_wrpll_funcs, DPLL_ID_WRPLL1,     0 },
+	{ "WRPLL 2",    &hsw_ddi_wrpll_funcs, DPLL_ID_WRPLL2,     0 },
+	{ "SPLL",       &hsw_ddi_spll_funcs,  DPLL_ID_SPLL,       0 },
+	{ "LCPLL 810",  &hsw_ddi_lcpll_funcs, DPLL_ID_LCPLL_810,  INTEL_DPLL_ALWAYS_ON },
+	{ "LCPLL 1350", &hsw_ddi_lcpll_funcs, DPLL_ID_LCPLL_1350, INTEL_DPLL_ALWAYS_ON },
+	{ "LCPLL 2700", &hsw_ddi_lcpll_funcs, DPLL_ID_LCPLL_2700, INTEL_DPLL_ALWAYS_ON },
+	{ },
+};
+
+static const struct intel_dpll_mgr hsw_pll_mgr = {
+	.dpll_info = hsw_plls,
+	.get_dplls = hsw_get_dpll,
+	.put_dplls = intel_put_dpll,
+	.update_ref_clks = hsw_update_dpll_ref_clks,
+	.dump_hw_state = hsw_dump_hw_state,
 };
 
 struct skl_dpll_regs {
@@ -1230,6 +1386,7 @@ struct skl_wrpll_params {
 
 static void skl_wrpll_params_populate(struct skl_wrpll_params *params,
 				      u64 afe_clock,
+				      int ref_clock,
 				      u64 central_freq,
 				      u32 p0, u32 p1, u32 p2)
 {
@@ -1289,14 +1446,15 @@ static void skl_wrpll_params_populate(struct skl_wrpll_params *params,
 	 * Intermediate values are in Hz.
 	 * Divide by MHz to match bsepc
 	 */
-	params->dco_integer = div_u64(dco_freq, 24 * MHz(1));
+	params->dco_integer = div_u64(dco_freq, ref_clock * KHz(1));
 	params->dco_fraction =
-		div_u64((div_u64(dco_freq, 24) -
+		div_u64((div_u64(dco_freq, ref_clock / KHz(1)) -
 			 params->dco_integer * MHz(1)) * 0x8000, MHz(1));
 }
 
 static bool
 skl_ddi_calculate_wrpll(int clock /* in Hz */,
+			int ref_clock,
 			struct skl_wrpll_params *wrpll_params)
 {
 	u64 afe_clock = clock * 5; /* AFE Clock is 5x Pixel clock */
@@ -1362,14 +1520,15 @@ skip_remaining_dividers:
 	 */
 	p0 = p1 = p2 = 0;
 	skl_wrpll_get_multipliers(ctx.p, &p0, &p1, &p2);
-	skl_wrpll_params_populate(wrpll_params, afe_clock, ctx.central_freq,
-				  p0, p1, p2);
+	skl_wrpll_params_populate(wrpll_params, afe_clock, ref_clock,
+				  ctx.central_freq, p0, p1, p2);
 
 	return true;
 }
 
 static bool skl_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state)
 {
+	struct drm_i915_private *i915 = to_i915(crtc_state->uapi.crtc->dev);
 	u32 ctrl1, cfgcr1, cfgcr2;
 	struct skl_wrpll_params wrpll_params = { 0, };
 
@@ -1382,6 +1541,7 @@ static bool skl_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state)
 	ctrl1 |= DPLL_CTRL1_HDMI_MODE(0);
 
 	if (!skl_ddi_calculate_wrpll(crtc_state->port_clock * 1000,
+				     i915->dpll.ref_clks.nssc,
 				     &wrpll_params))
 		return false;
 
@@ -1404,6 +1564,64 @@ static bool skl_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state)
 	return true;
 }
 
+static int skl_ddi_wrpll_get_freq(struct drm_i915_private *i915,
+				  const struct intel_shared_dpll *pll)
+{
+	const struct intel_dpll_hw_state *pll_state = &pll->state.hw_state;
+	int ref_clock = i915->dpll.ref_clks.nssc;
+	u32 p0, p1, p2, dco_freq;
+
+	p0 = pll_state->cfgcr2 & DPLL_CFGCR2_PDIV_MASK;
+	p2 = pll_state->cfgcr2 & DPLL_CFGCR2_KDIV_MASK;
+
+	if (pll_state->cfgcr2 &  DPLL_CFGCR2_QDIV_MODE(1))
+		p1 = (pll_state->cfgcr2 & DPLL_CFGCR2_QDIV_RATIO_MASK) >> 8;
+	else
+		p1 = 1;
+
+
+	switch (p0) {
+	case DPLL_CFGCR2_PDIV_1:
+		p0 = 1;
+		break;
+	case DPLL_CFGCR2_PDIV_2:
+		p0 = 2;
+		break;
+	case DPLL_CFGCR2_PDIV_3:
+		p0 = 3;
+		break;
+	case DPLL_CFGCR2_PDIV_7:
+		p0 = 7;
+		break;
+	}
+
+	switch (p2) {
+	case DPLL_CFGCR2_KDIV_5:
+		p2 = 5;
+		break;
+	case DPLL_CFGCR2_KDIV_2:
+		p2 = 2;
+		break;
+	case DPLL_CFGCR2_KDIV_3:
+		p2 = 3;
+		break;
+	case DPLL_CFGCR2_KDIV_1:
+		p2 = 1;
+		break;
+	}
+
+	dco_freq = (pll_state->cfgcr1 & DPLL_CFGCR1_DCO_INTEGER_MASK) *
+		   ref_clock;
+
+	dco_freq += ((pll_state->cfgcr1 & DPLL_CFGCR1_DCO_FRACTION_MASK) >> 9) *
+		    ref_clock / 0x8000;
+
+	if (WARN_ON(p0 == 0 || p1 == 0 || p2 == 0))
+		return 0;
+
+	return dco_freq / (p0 * p1 * p2 * 5);
+}
+
 static bool
 skl_ddi_dp_set_dpll_hw_state(struct intel_crtc_state *crtc_state)
 {
@@ -1444,6 +1662,40 @@ skl_ddi_dp_set_dpll_hw_state(struct intel_crtc_state *crtc_state)
 	return true;
 }
 
+static int skl_ddi_lcpll_get_freq(struct drm_i915_private *i915,
+				  const struct intel_shared_dpll *pll)
+{
+	int link_clock = 0;
+
+	switch ((pll->state.hw_state.ctrl1 &
+		 DPLL_CTRL1_LINK_RATE_MASK(0)) >>
+		DPLL_CTRL1_LINK_RATE_SHIFT(0)) {
+	case DPLL_CTRL1_LINK_RATE_810:
+		link_clock = 81000;
+		break;
+	case DPLL_CTRL1_LINK_RATE_1080:
+		link_clock = 108000;
+		break;
+	case DPLL_CTRL1_LINK_RATE_1350:
+		link_clock = 135000;
+		break;
+	case DPLL_CTRL1_LINK_RATE_1620:
+		link_clock = 162000;
+		break;
+	case DPLL_CTRL1_LINK_RATE_2160:
+		link_clock = 216000;
+		break;
+	case DPLL_CTRL1_LINK_RATE_2700:
+		link_clock = 270000;
+		break;
+	default:
+		drm_WARN(&i915->drm, 1, "Unsupported link rate\n");
+		break;
+	}
+
+	return link_clock * 2;
+}
+
 static bool skl_get_dpll(struct intel_atomic_state *state,
 			 struct intel_crtc *crtc,
 			 struct intel_encoder *encoder)
@@ -1493,6 +1745,25 @@ static bool skl_get_dpll(struct intel_atomic_state *state,
 	return true;
 }
 
+static int skl_ddi_pll_get_freq(struct drm_i915_private *i915,
+				const struct intel_shared_dpll *pll)
+{
+	/*
+	 * ctrl1 register is already shifted for each pll, just use 0 to get
+	 * the internal shift for each field
+	 */
+	if (pll->state.hw_state.ctrl1 & DPLL_CTRL1_HDMI_MODE(0))
+		return skl_ddi_wrpll_get_freq(i915, pll);
+	else
+		return skl_ddi_lcpll_get_freq(i915, pll);
+}
+
+static void skl_update_dpll_ref_clks(struct drm_i915_private *i915)
+{
+	/* No SSC ref */
+	i915->dpll.ref_clks.nssc = i915->cdclk.hw.ref;
+}
+
 static void skl_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state)
 {
@@ -1507,12 +1778,30 @@ static const struct intel_shared_dpll_funcs skl_ddi_pll_funcs = {
 	.enable = skl_ddi_pll_enable,
 	.disable = skl_ddi_pll_disable,
 	.get_hw_state = skl_ddi_pll_get_hw_state,
+	.get_freq = skl_ddi_pll_get_freq,
 };
 
 static const struct intel_shared_dpll_funcs skl_ddi_dpll0_funcs = {
 	.enable = skl_ddi_dpll0_enable,
 	.disable = skl_ddi_dpll0_disable,
 	.get_hw_state = skl_ddi_dpll0_get_hw_state,
+	.get_freq = skl_ddi_pll_get_freq,
+};
+
+static const struct dpll_info skl_plls[] = {
+	{ "DPLL 0", &skl_ddi_dpll0_funcs, DPLL_ID_SKL_DPLL0, INTEL_DPLL_ALWAYS_ON },
+	{ "DPLL 1", &skl_ddi_pll_funcs,   DPLL_ID_SKL_DPLL1, 0 },
+	{ "DPLL 2", &skl_ddi_pll_funcs,   DPLL_ID_SKL_DPLL2, 0 },
+	{ "DPLL 3", &skl_ddi_pll_funcs,   DPLL_ID_SKL_DPLL3, 0 },
+	{ },
+};
+
+static const struct intel_dpll_mgr skl_pll_mgr = {
+	.dpll_info = skl_plls,
+	.get_dplls = skl_get_dpll,
+	.put_dplls = intel_put_dpll,
+	.update_ref_clks = skl_update_dpll_ref_clks,
+	.dump_hw_state = skl_dump_hw_state,
 };
 
 static void bxt_ddi_pll_enable(struct drm_i915_private *dev_priv,
@@ -1903,6 +2192,23 @@ bxt_ddi_hdmi_set_dpll_hw_state(struct intel_crtc_state *crtc_state)
 	return bxt_ddi_set_dpll_hw_state(crtc_state, &clk_div);
 }
 
+static int bxt_ddi_pll_get_freq(struct drm_i915_private *i915,
+				const struct intel_shared_dpll *pll)
+{
+	const struct intel_dpll_hw_state *pll_state = &pll->state.hw_state;
+	struct dpll clock;
+
+	clock.m1 = 2;
+	clock.m2 = (pll_state->pll0 & PORT_PLL_M2_MASK) << 22;
+	if (pll_state->pll3 & PORT_PLL_M2_FRAC_ENABLE)
+		clock.m2 |= pll_state->pll2 & PORT_PLL_M2_FRAC_MASK;
+	clock.n = (pll_state->pll1 & PORT_PLL_N_MASK) >> PORT_PLL_N_SHIFT;
+	clock.p1 = (pll_state->ebb0 & PORT_PLL_P1_MASK) >> PORT_PLL_P1_SHIFT;
+	clock.p2 = (pll_state->ebb0 & PORT_PLL_P2_MASK) >> PORT_PLL_P2_SHIFT;
+
+	return chv_calc_dpll_params(i915->dpll.ref_clks.nssc, &clock);
+}
+
 static bool bxt_get_dpll(struct intel_atomic_state *state,
 			 struct intel_crtc *crtc,
 			 struct intel_encoder *encoder)
@@ -1936,6 +2242,13 @@ static bool bxt_get_dpll(struct intel_atomic_state *state,
 	return true;
 }
 
+static void bxt_update_dpll_ref_clks(struct drm_i915_private *i915)
+{
+	i915->dpll.ref_clks.ssc = 100000;
+	i915->dpll.ref_clks.nssc = 100000;
+	/* DSI non-SSC ref 19.2MHz */
+}
+
 static void bxt_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state)
 {
@@ -1959,66 +2272,7 @@ static const struct intel_shared_dpll_funcs bxt_ddi_pll_funcs = {
 	.enable = bxt_ddi_pll_enable,
 	.disable = bxt_ddi_pll_disable,
 	.get_hw_state = bxt_ddi_pll_get_hw_state,
-};
-
-struct intel_dpll_mgr {
-	const struct dpll_info *dpll_info;
-
-	bool (*get_dplls)(struct intel_atomic_state *state,
-			  struct intel_crtc *crtc,
-			  struct intel_encoder *encoder);
-	void (*put_dplls)(struct intel_atomic_state *state,
-			  struct intel_crtc *crtc);
-	void (*update_active_dpll)(struct intel_atomic_state *state,
-				   struct intel_crtc *crtc,
-				   struct intel_encoder *encoder);
-	void (*dump_hw_state)(struct drm_i915_private *dev_priv,
-			      const struct intel_dpll_hw_state *hw_state);
-};
-
-static const struct dpll_info pch_plls[] = {
-	{ "PCH DPLL A", &ibx_pch_dpll_funcs, DPLL_ID_PCH_PLL_A, 0 },
-	{ "PCH DPLL B", &ibx_pch_dpll_funcs, DPLL_ID_PCH_PLL_B, 0 },
-	{ },
-};
-
-static const struct intel_dpll_mgr pch_pll_mgr = {
-	.dpll_info = pch_plls,
-	.get_dplls = ibx_get_dpll,
-	.put_dplls = intel_put_dpll,
-	.dump_hw_state = ibx_dump_hw_state,
-};
-
-static const struct dpll_info hsw_plls[] = {
-	{ "WRPLL 1",    &hsw_ddi_wrpll_funcs, DPLL_ID_WRPLL1,     0 },
-	{ "WRPLL 2",    &hsw_ddi_wrpll_funcs, DPLL_ID_WRPLL2,     0 },
-	{ "SPLL",       &hsw_ddi_spll_funcs,  DPLL_ID_SPLL,       0 },
-	{ "LCPLL 810",  &hsw_ddi_lcpll_funcs, DPLL_ID_LCPLL_810,  INTEL_DPLL_ALWAYS_ON },
-	{ "LCPLL 1350", &hsw_ddi_lcpll_funcs, DPLL_ID_LCPLL_1350, INTEL_DPLL_ALWAYS_ON },
-	{ "LCPLL 2700", &hsw_ddi_lcpll_funcs, DPLL_ID_LCPLL_2700, INTEL_DPLL_ALWAYS_ON },
-	{ },
-};
-
-static const struct intel_dpll_mgr hsw_pll_mgr = {
-	.dpll_info = hsw_plls,
-	.get_dplls = hsw_get_dpll,
-	.put_dplls = intel_put_dpll,
-	.dump_hw_state = hsw_dump_hw_state,
-};
-
-static const struct dpll_info skl_plls[] = {
-	{ "DPLL 0", &skl_ddi_dpll0_funcs, DPLL_ID_SKL_DPLL0, INTEL_DPLL_ALWAYS_ON },
-	{ "DPLL 1", &skl_ddi_pll_funcs,   DPLL_ID_SKL_DPLL1, 0 },
-	{ "DPLL 2", &skl_ddi_pll_funcs,   DPLL_ID_SKL_DPLL2, 0 },
-	{ "DPLL 3", &skl_ddi_pll_funcs,   DPLL_ID_SKL_DPLL3, 0 },
-	{ },
-};
-
-static const struct intel_dpll_mgr skl_pll_mgr = {
-	.dpll_info = skl_plls,
-	.get_dplls = skl_get_dpll,
-	.put_dplls = intel_put_dpll,
-	.dump_hw_state = skl_dump_hw_state,
+	.get_freq = bxt_ddi_pll_get_freq,
 };
 
 static const struct dpll_info bxt_plls[] = {
@@ -2032,6 +2286,7 @@ static const struct intel_dpll_mgr bxt_pll_mgr = {
 	.dpll_info = bxt_plls,
 	.get_dplls = bxt_get_dpll,
 	.put_dplls = intel_put_dpll,
+	.update_ref_clks = bxt_update_dpll_ref_clks,
 	.dump_hw_state = bxt_dump_hw_state,
 };
 
@@ -2275,27 +2530,12 @@ static void cnl_wrpll_params_populate(struct skl_wrpll_params *params,
 	params->dco_fraction = dco & 0x7fff;
 }
 
-int cnl_hdmi_pll_ref_clock(struct drm_i915_private *dev_priv)
-{
-	int ref_clock = dev_priv->cdclk.hw.ref;
-
-	/*
-	 * For ICL+, the spec states: if reference frequency is 38.4,
-	 * use 19.2 because the DPLL automatically divides that by 2.
-	 */
-	if (INTEL_GEN(dev_priv) >= 11 && ref_clock == 38400)
-		ref_clock = 19200;
-
-	return ref_clock;
-}
-
 static bool
-cnl_ddi_calculate_wrpll(struct intel_crtc_state *crtc_state,
-			struct skl_wrpll_params *wrpll_params)
+__cnl_ddi_calculate_wrpll(struct intel_crtc_state *crtc_state,
+			  struct skl_wrpll_params *wrpll_params,
+			  int ref_clock)
 {
-	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;
 	u32 dco_max = 10000000;
 	u32 dco_mid = (dco_min + dco_max) / 2;
@@ -2327,15 +2567,22 @@ cnl_ddi_calculate_wrpll(struct intel_crtc_state *crtc_state,
 		return false;
 
 	cnl_wrpll_get_multipliers(best_div, &pdiv, &qdiv, &kdiv);
-
-	ref_clock = cnl_hdmi_pll_ref_clock(dev_priv);
-
 	cnl_wrpll_params_populate(wrpll_params, best_dco, ref_clock,
 				  pdiv, qdiv, kdiv);
 
 	return true;
 }
 
+static bool
+cnl_ddi_calculate_wrpll(struct intel_crtc_state *crtc_state,
+			struct skl_wrpll_params *wrpll_params)
+{
+	struct drm_i915_private *i915 = to_i915(crtc_state->uapi.crtc->dev);
+
+	return __cnl_ddi_calculate_wrpll(crtc_state, wrpll_params,
+					 i915->dpll.ref_clks.nssc);
+}
+
 static bool cnl_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state)
 {
 	u32 cfgcr0, cfgcr1;
@@ -2363,6 +2610,68 @@ static bool cnl_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state)
 	return true;
 }
 
+static int __cnl_ddi_wrpll_get_freq(struct drm_i915_private *dev_priv,
+				    const struct intel_shared_dpll *pll,
+				    int ref_clock)
+{
+	const struct intel_dpll_hw_state *pll_state = &pll->state.hw_state;
+	u32 p0, p1, p2, dco_freq;
+
+	p0 = pll_state->cfgcr1 & DPLL_CFGCR1_PDIV_MASK;
+	p2 = pll_state->cfgcr1 & DPLL_CFGCR1_KDIV_MASK;
+
+	if (pll_state->cfgcr1 & DPLL_CFGCR1_QDIV_MODE(1))
+		p1 = (pll_state->cfgcr1 & DPLL_CFGCR1_QDIV_RATIO_MASK) >>
+			DPLL_CFGCR1_QDIV_RATIO_SHIFT;
+	else
+		p1 = 1;
+
+
+	switch (p0) {
+	case DPLL_CFGCR1_PDIV_2:
+		p0 = 2;
+		break;
+	case DPLL_CFGCR1_PDIV_3:
+		p0 = 3;
+		break;
+	case DPLL_CFGCR1_PDIV_5:
+		p0 = 5;
+		break;
+	case DPLL_CFGCR1_PDIV_7:
+		p0 = 7;
+		break;
+	}
+
+	switch (p2) {
+	case DPLL_CFGCR1_KDIV_1:
+		p2 = 1;
+		break;
+	case DPLL_CFGCR1_KDIV_2:
+		p2 = 2;
+		break;
+	case DPLL_CFGCR1_KDIV_3:
+		p2 = 3;
+		break;
+	}
+
+	dco_freq = (pll_state->cfgcr0 & DPLL_CFGCR0_DCO_INTEGER_MASK) *
+		   ref_clock;
+
+	dco_freq += (((pll_state->cfgcr0 & DPLL_CFGCR0_DCO_FRACTION_MASK) >>
+		      DPLL_CFGCR0_DCO_FRACTION_SHIFT) * ref_clock) / 0x8000;
+
+	if (drm_WARN_ON(&dev_priv->drm, p0 == 0 || p1 == 0 || p2 == 0))
+		return 0;
+
+	return dco_freq / (p0 * p1 * p2 * 5);
+}
+
+static int cnl_ddi_wrpll_get_freq(struct drm_i915_private *i915,
+				  const struct intel_shared_dpll *pll)
+{
+	return __cnl_ddi_wrpll_get_freq(i915, pll, i915->dpll.ref_clks.nssc);
+}
+
 static bool
 cnl_ddi_dp_set_dpll_hw_state(struct intel_crtc_state *crtc_state)
 {
@@ -2408,6 +2717,44 @@ cnl_ddi_dp_set_dpll_hw_state(struct intel_crtc_state *crtc_state)
 	return true;
 }
 
+static int cnl_ddi_lcpll_get_freq(struct drm_i915_private *i915,
+				  const struct intel_shared_dpll *pll)
+{
+	int link_clock = 0;
+
+	switch (pll->state.hw_state.cfgcr0 & DPLL_CFGCR0_LINK_RATE_MASK) {
+	case DPLL_CFGCR0_LINK_RATE_810:
+		link_clock = 81000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_1080:
+		link_clock = 108000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_1350:
+		link_clock = 135000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_1620:
+		link_clock = 162000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_2160:
+		link_clock = 216000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_2700:
+		link_clock = 270000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_3240:
+		link_clock = 324000;
+		break;
+	case DPLL_CFGCR0_LINK_RATE_4050:
+		link_clock = 405000;
+		break;
+	default:
+		drm_WARN(&i915->drm, 1, "Unsupported link rate\n");
+		break;
+	}
+
+	return link_clock * 2;
+}
+
 static bool cnl_get_dpll(struct intel_atomic_state *state,
 			 struct intel_crtc *crtc,
 			 struct intel_encoder *encoder)
@@ -2457,6 +2804,21 @@ static bool cnl_get_dpll(struct intel_atomic_state *state,
 	return true;
 }
 
+static int cnl_ddi_pll_get_freq(struct drm_i915_private *i915,
+				const struct intel_shared_dpll *pll)
+{
+	if (pll->state.hw_state.cfgcr0 & DPLL_CFGCR0_HDMI_MODE)
+		return cnl_ddi_wrpll_get_freq(i915, pll);
+	else
+		return cnl_ddi_lcpll_get_freq(i915, pll);
+}
+
+static void cnl_update_dpll_ref_clks(struct drm_i915_private *i915)
+{
+	/* No SSC reference */
+	i915->dpll.ref_clks.nssc = i915->cdclk.hw.ref;
+}
+
 static void cnl_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state)
 {
@@ -2470,6 +2832,7 @@ static const struct intel_shared_dpll_funcs cnl_ddi_pll_funcs = {
 	.enable = cnl_ddi_pll_enable,
 	.disable = cnl_ddi_pll_disable,
 	.get_hw_state = cnl_ddi_pll_get_hw_state,
+	.get_freq = cnl_ddi_pll_get_freq,
 };
 
 static const struct dpll_info cnl_plls[] = {
@@ -2483,6 +2846,7 @@ static const struct intel_dpll_mgr cnl_pll_mgr = {
 	.dpll_info = cnl_plls,
 	.get_dplls = cnl_get_dpll,
 	.put_dplls = intel_put_dpll,
+	.update_ref_clks = cnl_update_dpll_ref_clks,
 	.dump_hw_state = cnl_dump_hw_state,
 };
 
@@ -2578,7 +2942,7 @@ static bool icl_calc_dp_combo_pll(struct intel_crtc_state *crtc_state,
 {
 	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 ?
+		dev_priv->dpll.ref_clks.nssc == 24000 ?
 		icl_dp_combo_pll_24MHz_values :
 		icl_dp_combo_pll_19_2MHz_values;
 	int clock = crtc_state->port_clock;
@@ -2601,9 +2965,9 @@ static bool icl_calc_tbt_pll(struct intel_crtc_state *crtc_state,
 	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) {
+		switch (dev_priv->dpll.ref_clks.nssc) {
 		default:
-			MISSING_CASE(dev_priv->cdclk.hw.ref);
+			MISSING_CASE(dev_priv->dpll.ref_clks.nssc);
 			/* fall-through */
 		case 19200:
 		case 38400:
@@ -2614,9 +2978,9 @@ static bool icl_calc_tbt_pll(struct intel_crtc_state *crtc_state,
 			break;
 		}
 	} else {
-		switch (dev_priv->cdclk.hw.ref) {
+		switch (dev_priv->dpll.ref_clks.nssc) {
 		default:
-			MISSING_CASE(dev_priv->cdclk.hw.ref);
+			MISSING_CASE(dev_priv->dpll.ref_clks.nssc);
 			/* fall-through */
 		case 19200:
 		case 38400:
@@ -2631,6 +2995,49 @@ static bool icl_calc_tbt_pll(struct intel_crtc_state *crtc_state,
 	return true;
 }
 
+static int icl_ddi_tbt_pll_get_freq(struct drm_i915_private *i915,
+				    const struct intel_shared_dpll *pll)
+{
+	/*
+	 * The PLL outputs multiple frequencies at the same time, selection is
+	 * made at DDI clock mux level.
+	 */
+	drm_WARN_ON(&i915->drm, 1);
+
+	return 0;
+}
+
+static int icl_wrpll_ref_clock(struct drm_i915_private *i915)
+{
+	int ref_clock = i915->dpll.ref_clks.nssc;
+
+	/*
+	 * For ICL+, the spec states: if reference frequency is 38.4,
+	 * use 19.2 because the DPLL automatically divides that by 2.
+	 */
+	if (ref_clock == 38400)
+		ref_clock = 19200;
+
+	return ref_clock;
+}
+
+static bool
+icl_calc_wrpll(struct intel_crtc_state *crtc_state,
+	       struct skl_wrpll_params *wrpll_params)
+{
+	struct drm_i915_private *i915 = to_i915(crtc_state->uapi.crtc->dev);
+
+	return __cnl_ddi_calculate_wrpll(crtc_state, wrpll_params,
+					 icl_wrpll_ref_clock(i915));
+}
+
+static int icl_ddi_combo_pll_get_freq(struct drm_i915_private *i915,
+				      const struct intel_shared_dpll *pll)
+{
+	return __cnl_ddi_wrpll_get_freq(i915, pll,
+					icl_wrpll_ref_clock(i915));
+}
+
 static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
 				struct intel_encoder *encoder,
 				struct intel_dpll_hw_state *pll_state)
@@ -2645,7 +3052,7 @@ static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
 		ret = icl_calc_tbt_pll(crtc_state, &pll_params);
 	else if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI) ||
 		 intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI))
-		ret = cnl_ddi_calculate_wrpll(crtc_state, &pll_params);
+		ret = icl_calc_wrpll(crtc_state, &pll_params);
 	else
 		ret = icl_calc_dp_combo_pll(crtc_state, &pll_params);
 
@@ -2768,7 +3175,7 @@ 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->uapi.crtc->dev);
-	int refclk_khz = dev_priv->cdclk.hw.ref;
+	int refclk_khz = dev_priv->dpll.ref_clks.nssc;
 	int clock = crtc_state->port_clock;
 	u32 dco_khz, m1div, m2div_int, m2div_rem, m2div_frac;
 	u32 iref_ndiv, iref_trim, iref_pulse_w;
@@ -2969,6 +3376,78 @@ static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
 	return true;
 }
 
+static int icl_ddi_mg_pll_get_freq(struct drm_i915_private *dev_priv,
+				   const struct intel_shared_dpll *pll)
+{
+	const struct intel_dpll_hw_state *pll_state = &pll->state.hw_state;
+	u32 m1, m2_int, m2_frac, div1, div2, ref_clock;
+	u64 tmp;
+
+	ref_clock = dev_priv->dpll.ref_clks.nssc;
+
+	if (INTEL_GEN(dev_priv) >= 12) {
+		m1 = pll_state->mg_pll_div0 & DKL_PLL_DIV0_FBPREDIV_MASK;
+		m1 = m1 >> DKL_PLL_DIV0_FBPREDIV_SHIFT;
+		m2_int = pll_state->mg_pll_div0 & DKL_PLL_DIV0_FBDIV_INT_MASK;
+
+		if (pll_state->mg_pll_bias & DKL_PLL_BIAS_FRAC_EN_H) {
+			m2_frac = pll_state->mg_pll_bias &
+				  DKL_PLL_BIAS_FBDIV_FRAC_MASK;
+			m2_frac = m2_frac >> DKL_PLL_BIAS_FBDIV_SHIFT;
+		} else {
+			m2_frac = 0;
+		}
+	} else {
+		m1 = pll_state->mg_pll_div1 & MG_PLL_DIV1_FBPREDIV_MASK;
+		m2_int = pll_state->mg_pll_div0 & MG_PLL_DIV0_FBDIV_INT_MASK;
+
+		if (pll_state->mg_pll_div0 & MG_PLL_DIV0_FRACNEN_H) {
+			m2_frac = pll_state->mg_pll_div0 &
+				  MG_PLL_DIV0_FBDIV_FRAC_MASK;
+			m2_frac = m2_frac >> MG_PLL_DIV0_FBDIV_FRAC_SHIFT;
+		} else {
+			m2_frac = 0;
+		}
+	}
+
+	switch (pll_state->mg_clktop2_hsclkctl &
+		MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_MASK) {
+	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_2:
+		div1 = 2;
+		break;
+	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_3:
+		div1 = 3;
+		break;
+	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_5:
+		div1 = 5;
+		break;
+	case MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO_7:
+		div1 = 7;
+		break;
+	default:
+		MISSING_CASE(pll_state->mg_clktop2_hsclkctl);
+		return 0;
+	}
+
+	div2 = (pll_state->mg_clktop2_hsclkctl &
+		MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO_MASK) >>
+		MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO_SHIFT;
+
+	/* div2 value of 0 is same as 1 means no div */
+	if (div2 == 0)
+		div2 = 1;
+
+	/*
+	 * Adjust the original formula to delay the division by 2^22 in order to
+	 * minimize possible rounding errors.
+	 */
+	tmp = (u64)m1 * m2_int * ref_clock +
+	      (((u64)m1 * m2_frac * ref_clock) >> 22);
+	tmp = div_u64(tmp, 5 * div1 * div2);
+
+	return tmp;
+}
+
 /**
  * icl_set_active_port_dpll - select the active port DPLL for a given CRTC
  * @crtc_state: state for the CRTC to select the DPLL for
@@ -3201,7 +3680,7 @@ static bool mg_pll_get_hw_state(struct drm_i915_private *dev_priv,
 	hw_state->mg_pll_tdc_coldst_bias =
 		intel_de_read(dev_priv, MG_PLL_TDC_COLDST_BIAS(tc_port));
 
-	if (dev_priv->cdclk.hw.ref == 38400) {
+	if (dev_priv->dpll.ref_clks.nssc == 38400) {
 		hw_state->mg_pll_tdc_coldst_bias_mask = MG_PLL_TDC_COLDST_COLDSTART;
 		hw_state->mg_pll_bias_mask = 0;
 	} else {
@@ -3682,6 +4161,12 @@ static void mg_pll_disable(struct drm_i915_private *dev_priv,
 	icl_pll_disable(dev_priv, pll, enable_reg);
 }
 
+static void icl_update_dpll_ref_clks(struct drm_i915_private *i915)
+{
+	/* No SSC ref */
+	i915->dpll.ref_clks.nssc = i915->cdclk.hw.ref;
+}
+
 static void icl_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state)
 {
@@ -3709,18 +4194,21 @@ static const struct intel_shared_dpll_funcs combo_pll_funcs = {
 	.enable = combo_pll_enable,
 	.disable = combo_pll_disable,
 	.get_hw_state = combo_pll_get_hw_state,
+	.get_freq = icl_ddi_combo_pll_get_freq,
 };
 
 static const struct intel_shared_dpll_funcs tbt_pll_funcs = {
 	.enable = tbt_pll_enable,
 	.disable = tbt_pll_disable,
 	.get_hw_state = tbt_pll_get_hw_state,
+	.get_freq = icl_ddi_tbt_pll_get_freq,
 };
 
 static const struct intel_shared_dpll_funcs mg_pll_funcs = {
 	.enable = mg_pll_enable,
 	.disable = mg_pll_disable,
 	.get_hw_state = mg_pll_get_hw_state,
+	.get_freq = icl_ddi_mg_pll_get_freq,
 };
 
 static const struct dpll_info icl_plls[] = {
@@ -3739,6 +4227,7 @@ static const struct intel_dpll_mgr icl_pll_mgr = {
 	.get_dplls = icl_get_dplls,
 	.put_dplls = icl_put_dplls,
 	.update_active_dpll = icl_update_active_dpll,
+	.update_ref_clks = icl_update_dpll_ref_clks,
 	.dump_hw_state = icl_dump_hw_state,
 };
 
@@ -3753,6 +4242,7 @@ static const struct intel_dpll_mgr ehl_pll_mgr = {
 	.dpll_info = ehl_plls,
 	.get_dplls = icl_get_dplls,
 	.put_dplls = icl_put_dplls,
+	.update_ref_clks = icl_update_dpll_ref_clks,
 	.dump_hw_state = icl_dump_hw_state,
 };
 
@@ -3760,6 +4250,7 @@ static const struct intel_shared_dpll_funcs dkl_pll_funcs = {
 	.enable = mg_pll_enable,
 	.disable = mg_pll_disable,
 	.get_hw_state = dkl_pll_get_hw_state,
+	.get_freq = icl_ddi_mg_pll_get_freq,
 };
 
 static const struct dpll_info tgl_plls[] = {
@@ -3780,6 +4271,7 @@ static const struct intel_dpll_mgr tgl_pll_mgr = {
 	.get_dplls = icl_get_dplls,
 	.put_dplls = icl_put_dplls,
 	.update_active_dpll = icl_update_active_dpll,
+	.update_ref_clks = icl_update_dpll_ref_clks,
 	.dump_hw_state = icl_dump_hw_state,
 };
 
@@ -3814,7 +4306,7 @@ void intel_shared_dpll_init(struct drm_device *dev)
 		dpll_mgr = &pch_pll_mgr;
 
 	if (!dpll_mgr) {
-		dev_priv->num_shared_dpll = 0;
+		dev_priv->dpll.num_shared_dpll = 0;
 		return;
 	}
 
@@ -3822,14 +4314,14 @@ void intel_shared_dpll_init(struct drm_device *dev)
 
 	for (i = 0; dpll_info[i].name; i++) {
 		drm_WARN_ON(dev, i != dpll_info[i].id);
-		dev_priv->shared_dplls[i].info = &dpll_info[i];
+		dev_priv->dpll.shared_dplls[i].info = &dpll_info[i];
 	}
 
-	dev_priv->dpll_mgr = dpll_mgr;
-	dev_priv->num_shared_dpll = i;
-	mutex_init(&dev_priv->dpll_lock);
+	dev_priv->dpll.mgr = dpll_mgr;
+	dev_priv->dpll.num_shared_dpll = i;
+	mutex_init(&dev_priv->dpll.lock);
 
-	BUG_ON(dev_priv->num_shared_dpll > I915_NUM_PLLS);
+	BUG_ON(dev_priv->dpll.num_shared_dpll > I915_NUM_PLLS);
 }
 
 /**
@@ -3856,7 +4348,7 @@ bool intel_reserve_shared_dplls(struct intel_atomic_state *state,
 				struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	const struct intel_dpll_mgr *dpll_mgr = dev_priv->dpll_mgr;
+	const struct intel_dpll_mgr *dpll_mgr = dev_priv->dpll.mgr;
 
 	if (drm_WARN_ON(&dev_priv->drm, !dpll_mgr))
 		return false;
@@ -3879,7 +4371,7 @@ void intel_release_shared_dplls(struct intel_atomic_state *state,
 				struct intel_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	const struct intel_dpll_mgr *dpll_mgr = dev_priv->dpll_mgr;
+	const struct intel_dpll_mgr *dpll_mgr = dev_priv->dpll.mgr;
 
 	/*
 	 * FIXME: this function is called for every platform having a
@@ -3908,7 +4400,7 @@ void intel_update_active_dpll(struct intel_atomic_state *state,
 			      struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	const struct intel_dpll_mgr *dpll_mgr = dev_priv->dpll_mgr;
+	const struct intel_dpll_mgr *dpll_mgr = dev_priv->dpll.mgr;
 
 	if (drm_WARN_ON(&dev_priv->drm, !dpll_mgr))
 		return;
@@ -3917,6 +4409,84 @@ void intel_update_active_dpll(struct intel_atomic_state *state,
 }
 
 /**
+ * intel_dpll_get_freq - calculate the DPLL's output frequency
+ * @i915: i915 device
+ * @pll: DPLL for which to calculate the output frequency
+ *
+ * Return the output frequency corresponding to @pll's current state.
+ */
+int intel_dpll_get_freq(struct drm_i915_private *i915,
+			const struct intel_shared_dpll *pll)
+{
+	if (drm_WARN_ON(&i915->drm, !pll->info->funcs->get_freq))
+		return 0;
+
+	return pll->info->funcs->get_freq(i915, pll);
+}
+
+static void readout_dpll_hw_state(struct drm_i915_private *i915,
+				  struct intel_shared_dpll *pll)
+{
+	struct intel_crtc *crtc;
+
+	pll->on = pll->info->funcs->get_hw_state(i915, pll,
+						 &pll->state.hw_state);
+
+	if (IS_ELKHARTLAKE(i915) && pll->on &&
+	    pll->info->id == DPLL_ID_EHL_DPLL4) {
+		pll->wakeref = intel_display_power_get(i915,
+						       POWER_DOMAIN_DPLL_DC_OFF);
+	}
+
+	pll->state.crtc_mask = 0;
+	for_each_intel_crtc(&i915->drm, crtc) {
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
+
+		if (crtc_state->hw.active && crtc_state->shared_dpll == pll)
+			pll->state.crtc_mask |= 1 << crtc->pipe;
+	}
+	pll->active_mask = pll->state.crtc_mask;
+
+	drm_dbg_kms(&i915->drm,
+		    "%s hw state readout: crtc_mask 0x%08x, on %i\n",
+		    pll->info->name, pll->state.crtc_mask, pll->on);
+}
+
+void intel_dpll_readout_hw_state(struct drm_i915_private *i915)
+{
+	int i;
+
+	if (i915->dpll.mgr && i915->dpll.mgr->update_ref_clks)
+		i915->dpll.mgr->update_ref_clks(i915);
+
+	for (i = 0; i < i915->dpll.num_shared_dpll; i++)
+		readout_dpll_hw_state(i915, &i915->dpll.shared_dplls[i]);
+}
+
+static void sanitize_dpll_state(struct drm_i915_private *i915,
+				struct intel_shared_dpll *pll)
+{
+	if (!pll->on || pll->active_mask)
+		return;
+
+	drm_dbg_kms(&i915->drm,
+		    "%s enabled but not in use, disabling\n",
+		    pll->info->name);
+
+	pll->info->funcs->disable(i915, pll);
+	pll->on = false;
+}
+
+void intel_dpll_sanitize_state(struct drm_i915_private *i915)
+{
+	int i;
+
+	for (i = 0; i < i915->dpll.num_shared_dpll; i++)
+		sanitize_dpll_state(i915, &i915->dpll.shared_dplls[i]);
+}
+
+/**
  * intel_shared_dpll_dump_hw_state - write hw_state to dmesg
  * @dev_priv: i915 drm device
  * @hw_state: hw state to be written to the log
@@ -3926,8 +4496,8 @@ void intel_update_active_dpll(struct intel_atomic_state *state,
 void intel_dpll_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state)
 {
-	if (dev_priv->dpll_mgr) {
-		dev_priv->dpll_mgr->dump_hw_state(dev_priv, hw_state);
+	if (dev_priv->dpll.mgr) {
+		dev_priv->dpll.mgr->dump_hw_state(dev_priv, hw_state);
 	} else {
 		/* fallback for platforms that don't use the shared dpll
 		 * infrastructure
diff --git a/drivers/gpu/drm/i915/display/intel_dpll_mgr.h b/drivers/gpu/drm/i915/display/intel_dpll_mgr.h
index 2a104c64291d..5d9a2bc371e7 100644
--- a/drivers/gpu/drm/i915/display/intel_dpll_mgr.h
+++ b/drivers/gpu/drm/i915/display/intel_dpll_mgr.h
@@ -278,6 +278,15 @@ struct intel_shared_dpll_funcs {
 	bool (*get_hw_state)(struct drm_i915_private *dev_priv,
 			     struct intel_shared_dpll *pll,
 			     struct intel_dpll_hw_state *hw_state);
+
+	/**
+	 * @get_freq:
+	 *
+	 * Hook for calculating the pll's output frequency based on its
+	 * current state.
+	 */
+	int (*get_freq)(struct drm_i915_private *i915,
+			const struct intel_shared_dpll *pll);
 };
 
 /**
@@ -372,15 +381,18 @@ void icl_set_active_port_dpll(struct intel_crtc_state *crtc_state,
 void intel_update_active_dpll(struct intel_atomic_state *state,
 			      struct intel_crtc *crtc,
 			      struct intel_encoder *encoder);
+int intel_dpll_get_freq(struct drm_i915_private *i915,
+			const struct intel_shared_dpll *pll);
 void intel_prepare_shared_dpll(const struct intel_crtc_state *crtc_state);
 void intel_enable_shared_dpll(const struct intel_crtc_state *crtc_state);
 void intel_disable_shared_dpll(const struct intel_crtc_state *crtc_state);
 void intel_shared_dpll_swap_state(struct intel_atomic_state *state);
 void intel_shared_dpll_init(struct drm_device *dev);
+void intel_dpll_readout_hw_state(struct drm_i915_private *dev_priv);
+void intel_dpll_sanitize_state(struct drm_i915_private *dev_priv);
 
 void intel_dpll_dump_hw_state(struct drm_i915_private *dev_priv,
 			      const struct intel_dpll_hw_state *hw_state);
-int cnl_hdmi_pll_ref_clock(struct drm_i915_private *dev_priv);
 enum intel_dpll_id icl_tc_port_to_pll_id(enum tc_port tc_port);
 bool intel_dpll_is_combophy(enum intel_dpll_id id);
 
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.c b/drivers/gpu/drm/i915/display/intel_dsb.c
index 76ae01277fd6..d7a6bf2277df 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.c
+++ b/drivers/gpu/drm/i915/display/intel_dsb.c
@@ -52,7 +52,7 @@ static inline bool intel_dsb_enable_engine(struct intel_dsb *dsb)
 
 	dsb_ctrl = intel_de_read(dev_priv, DSB_CTRL(pipe, dsb->id));
 	if (DSB_STATUS & dsb_ctrl) {
-		DRM_DEBUG_KMS("DSB engine is busy.\n");
+		drm_dbg_kms(&dev_priv->drm, "DSB engine is busy.\n");
 		return false;
 	}
 
@@ -72,7 +72,7 @@ static inline bool intel_dsb_disable_engine(struct intel_dsb *dsb)
 
 	dsb_ctrl = intel_de_read(dev_priv, DSB_CTRL(pipe, dsb->id));
 	if (DSB_STATUS & dsb_ctrl) {
-		DRM_DEBUG_KMS("DSB engine is busy.\n");
+		drm_dbg_kms(&dev_priv->drm, "DSB engine is busy.\n");
 		return false;
 	}
 
@@ -115,20 +115,20 @@ intel_dsb_get(struct intel_crtc *crtc)
 
 	obj = i915_gem_object_create_internal(i915, DSB_BUF_SIZE);
 	if (IS_ERR(obj)) {
-		DRM_ERROR("Gem object creation failed\n");
+		drm_err(&i915->drm, "Gem object creation failed\n");
 		goto out;
 	}
 
 	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
 	if (IS_ERR(vma)) {
-		DRM_ERROR("Vma creation failed\n");
+		drm_err(&i915->drm, "Vma creation failed\n");
 		i915_gem_object_put(obj);
 		goto out;
 	}
 
 	buf = i915_gem_object_pin_map(vma->obj, I915_MAP_WC);
 	if (IS_ERR(buf)) {
-		DRM_ERROR("Command buffer creation failed\n");
+		drm_err(&i915->drm, "Command buffer creation failed\n");
 		goto out;
 	}
 
@@ -203,7 +203,7 @@ void intel_dsb_indexed_reg_write(struct intel_dsb *dsb, i915_reg_t reg,
 	}
 
 	if (drm_WARN_ON(&dev_priv->drm, dsb->free_pos >= DSB_BUF_SIZE)) {
-		DRM_DEBUG_KMS("DSB buffer overflow\n");
+		drm_dbg_kms(&dev_priv->drm, "DSB buffer overflow\n");
 		return;
 	}
 
@@ -277,7 +277,7 @@ void intel_dsb_reg_write(struct intel_dsb *dsb, i915_reg_t reg, u32 val)
 	}
 
 	if (drm_WARN_ON(&dev_priv->drm, dsb->free_pos >= DSB_BUF_SIZE)) {
-		DRM_DEBUG_KMS("DSB buffer overflow\n");
+		drm_dbg_kms(&dev_priv->drm, "DSB buffer overflow\n");
 		return;
 	}
 
@@ -310,7 +310,8 @@ void intel_dsb_commit(struct intel_dsb *dsb)
 		goto reset;
 
 	if (is_dsb_busy(dsb)) {
-		DRM_ERROR("HEAD_PTR write failed - dsb engine is busy.\n");
+		drm_err(&dev_priv->drm,
+			"HEAD_PTR write failed - dsb engine is busy.\n");
 		goto reset;
 	}
 	intel_de_write(dev_priv, DSB_HEAD(pipe, dsb->id),
@@ -322,15 +323,18 @@ void intel_dsb_commit(struct intel_dsb *dsb)
 		       (tail - dsb->free_pos * 4));
 
 	if (is_dsb_busy(dsb)) {
-		DRM_ERROR("TAIL_PTR write failed - dsb engine is busy.\n");
+		drm_err(&dev_priv->drm,
+			"TAIL_PTR write failed - dsb engine is busy.\n");
 		goto reset;
 	}
-	DRM_DEBUG_KMS("DSB execution started - head 0x%x, tail 0x%x\n",
-		      i915_ggtt_offset(dsb->vma), tail);
+	drm_dbg_kms(&dev_priv->drm,
+		    "DSB execution started - head 0x%x, tail 0x%x\n",
+		    i915_ggtt_offset(dsb->vma), tail);
 	intel_de_write(dev_priv, DSB_TAIL(pipe, dsb->id),
 		       i915_ggtt_offset(dsb->vma) + tail);
 	if (wait_for(!is_dsb_busy(dsb), 1)) {
-		DRM_ERROR("Timed out waiting for DSB workload completion.\n");
+		drm_err(&dev_priv->drm,
+			"Timed out waiting for DSB workload completion.\n");
 		goto reset;
 	}
 
diff --git a/drivers/gpu/drm/i915/display/intel_dsi_vbt.c b/drivers/gpu/drm/i915/display/intel_dsi_vbt.c
index 694498f4b719..574dcfec9577 100644
--- a/drivers/gpu/drm/i915/display/intel_dsi_vbt.c
+++ b/drivers/gpu/drm/i915/display/intel_dsi_vbt.c
@@ -36,7 +36,6 @@
 
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
-#include <drm/i915_drm.h>
 
 #include <video/mipi_display.h>
 
diff --git a/drivers/gpu/drm/i915/display/intel_dvo.c b/drivers/gpu/drm/i915/display/intel_dvo.c
index 77f3d083b7a1..341d5ce8b062 100644
--- a/drivers/gpu/drm/i915/display/intel_dvo.c
+++ b/drivers/gpu/drm/i915/display/intel_dvo.c
@@ -30,7 +30,6 @@
 
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_connector.h"
diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c
index ddf8d3bb7a7d..2e5d835a9eaa 100644
--- a/drivers/gpu/drm/i915/display/intel_fbc.c
+++ b/drivers/gpu/drm/i915/display/intel_fbc.c
@@ -42,6 +42,7 @@
 
 #include "i915_drv.h"
 #include "i915_trace.h"
+#include "i915_vgpu.h"
 #include "intel_display_types.h"
 #include "intel_fbc.h"
 #include "intel_frontbuffer.h"
@@ -320,7 +321,7 @@ static void gen7_fbc_activate(struct drm_i915_private *dev_priv)
 			       SNB_CPU_FENCE_ENABLE | params->fence_id);
 		intel_de_write(dev_priv, DPFC_CPU_FENCE_OFFSET,
 			       params->crtc.fence_y_offset);
-	} else {
+	} else if (dev_priv->ggtt.num_fences) {
 		intel_de_write(dev_priv, SNB_DPFC_CTL_SA, 0);
 		intel_de_write(dev_priv, DPFC_CPU_FENCE_OFFSET, 0);
 	}
@@ -508,12 +509,12 @@ static int intel_fbc_alloc_cfb(struct drm_i915_private *dev_priv,
 
 		fbc->compressed_llb = compressed_llb;
 
-		GEM_BUG_ON(range_overflows_t(u64, dev_priv->dsm.start,
-					     fbc->compressed_fb.start,
-					     U32_MAX));
-		GEM_BUG_ON(range_overflows_t(u64, dev_priv->dsm.start,
-					     fbc->compressed_llb->start,
-					     U32_MAX));
+		GEM_BUG_ON(range_overflows_end_t(u64, dev_priv->dsm.start,
+						 fbc->compressed_fb.start,
+						 U32_MAX));
+		GEM_BUG_ON(range_overflows_end_t(u64, dev_priv->dsm.start,
+						 fbc->compressed_llb->start,
+						 U32_MAX));
 		intel_de_write(dev_priv, FBC_CFB_BASE,
 			       dev_priv->dsm.start + fbc->compressed_fb.start);
 		intel_de_write(dev_priv, FBC_LL_BASE,
@@ -691,12 +692,37 @@ static bool intel_fbc_cfb_size_changed(struct drm_i915_private *dev_priv)
 		fbc->compressed_fb.size * fbc->threshold;
 }
 
+static bool intel_fbc_can_enable(struct drm_i915_private *dev_priv)
+{
+	struct intel_fbc *fbc = &dev_priv->fbc;
+
+	if (intel_vgpu_active(dev_priv)) {
+		fbc->no_fbc_reason = "VGPU is active";
+		return false;
+	}
+
+	if (!i915_modparams.enable_fbc) {
+		fbc->no_fbc_reason = "disabled per module param or by default";
+		return false;
+	}
+
+	if (fbc->underrun_detected) {
+		fbc->no_fbc_reason = "underrun detected";
+		return false;
+	}
+
+	return true;
+}
+
 static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 {
 	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;
 
+	if (!intel_fbc_can_enable(dev_priv))
+		return false;
+
 	if (!cache->plane.visible) {
 		fbc->no_fbc_reason = "primary plane not visible";
 		return false;
@@ -795,28 +821,6 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 	return true;
 }
 
-static bool intel_fbc_can_enable(struct drm_i915_private *dev_priv)
-{
-	struct intel_fbc *fbc = &dev_priv->fbc;
-
-	if (intel_vgpu_active(dev_priv)) {
-		fbc->no_fbc_reason = "VGPU is active";
-		return false;
-	}
-
-	if (!i915_modparams.enable_fbc) {
-		fbc->no_fbc_reason = "disabled per module param or by default";
-		return false;
-	}
-
-	if (fbc->underrun_detected) {
-		fbc->no_fbc_reason = "underrun detected";
-		return false;
-	}
-
-	return true;
-}
-
 static void intel_fbc_get_reg_params(struct intel_crtc *crtc,
 				     struct intel_fbc_reg_params *params)
 {
diff --git a/drivers/gpu/drm/i915/display/intel_fbdev.c b/drivers/gpu/drm/i915/display/intel_fbdev.c
index 876264fc6560..3bc804212a99 100644
--- a/drivers/gpu/drm/i915/display/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/display/intel_fbdev.c
@@ -40,7 +40,6 @@
 #include <drm/drm_crtc.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_fourcc.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_display_types.h"
diff --git a/drivers/gpu/drm/i915/display/intel_fifo_underrun.c b/drivers/gpu/drm/i915/display/intel_fifo_underrun.c
index 470b3b0b9bdb..813a4f7033e1 100644
--- a/drivers/gpu/drm/i915/display/intel_fifo_underrun.c
+++ b/drivers/gpu/drm/i915/display/intel_fifo_underrun.c
@@ -103,7 +103,7 @@ static void i9xx_check_fifo_underruns(struct intel_crtc *crtc)
 	intel_de_posting_read(dev_priv, reg);
 
 	trace_intel_cpu_fifo_underrun(dev_priv, crtc->pipe);
-	DRM_ERROR("pipe %c underrun\n", pipe_name(crtc->pipe));
+	drm_err(&dev_priv->drm, "pipe %c underrun\n", pipe_name(crtc->pipe));
 }
 
 static void i9xx_set_fifo_underrun_reporting(struct drm_device *dev,
@@ -123,7 +123,8 @@ static void i9xx_set_fifo_underrun_reporting(struct drm_device *dev,
 		intel_de_posting_read(dev_priv, reg);
 	} else {
 		if (old && intel_de_read(dev_priv, reg) & PIPE_FIFO_UNDERRUN_STATUS)
-			DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+			drm_err(&dev_priv->drm, "pipe %c underrun\n",
+				pipe_name(pipe));
 	}
 }
 
@@ -155,7 +156,7 @@ static void ivb_check_fifo_underruns(struct intel_crtc *crtc)
 	intel_de_posting_read(dev_priv, GEN7_ERR_INT);
 
 	trace_intel_cpu_fifo_underrun(dev_priv, pipe);
-	DRM_ERROR("fifo underrun on pipe %c\n", pipe_name(pipe));
+	drm_err(&dev_priv->drm, "fifo underrun on pipe %c\n", pipe_name(pipe));
 }
 
 static void ivb_set_fifo_underrun_reporting(struct drm_device *dev,
@@ -176,8 +177,9 @@ static void ivb_set_fifo_underrun_reporting(struct drm_device *dev,
 
 		if (old &&
 		    intel_de_read(dev_priv, GEN7_ERR_INT) & ERR_INT_FIFO_UNDERRUN(pipe)) {
-			DRM_ERROR("uncleared fifo underrun on pipe %c\n",
-				  pipe_name(pipe));
+			drm_err(&dev_priv->drm,
+				"uncleared fifo underrun on pipe %c\n",
+				pipe_name(pipe));
 		}
 	}
 }
@@ -223,8 +225,8 @@ static void cpt_check_pch_fifo_underruns(struct intel_crtc *crtc)
 	intel_de_posting_read(dev_priv, SERR_INT);
 
 	trace_intel_pch_fifo_underrun(dev_priv, pch_transcoder);
-	DRM_ERROR("pch fifo underrun on pch transcoder %c\n",
-		  pipe_name(pch_transcoder));
+	drm_err(&dev_priv->drm, "pch fifo underrun on pch transcoder %c\n",
+		pipe_name(pch_transcoder));
 }
 
 static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
@@ -246,8 +248,9 @@ static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
 
 		if (old && intel_de_read(dev_priv, SERR_INT) &
 		    SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder)) {
-			DRM_ERROR("uncleared pch fifo underrun on pch transcoder %c\n",
-				  pipe_name(pch_transcoder));
+			drm_err(&dev_priv->drm,
+				"uncleared pch fifo underrun on pch transcoder %c\n",
+				pipe_name(pch_transcoder));
 		}
 	}
 }
@@ -381,8 +384,8 @@ void intel_cpu_fifo_underrun_irq_handler(struct drm_i915_private *dev_priv,
 
 	if (intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false)) {
 		trace_intel_cpu_fifo_underrun(dev_priv, pipe);
-		DRM_ERROR("CPU pipe %c FIFO underrun\n",
-			  pipe_name(pipe));
+		drm_err(&dev_priv->drm, "CPU pipe %c FIFO underrun\n",
+			pipe_name(pipe));
 	}
 
 	intel_fbc_handle_fifo_underrun_irq(dev_priv);
@@ -403,8 +406,8 @@ void intel_pch_fifo_underrun_irq_handler(struct drm_i915_private *dev_priv,
 	if (intel_set_pch_fifo_underrun_reporting(dev_priv, pch_transcoder,
 						  false)) {
 		trace_intel_pch_fifo_underrun(dev_priv, pch_transcoder);
-		DRM_ERROR("PCH transcoder %c FIFO underrun\n",
-			  pipe_name(pch_transcoder));
+		drm_err(&dev_priv->drm, "PCH transcoder %c FIFO underrun\n",
+			pipe_name(pch_transcoder));
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/display/intel_gmbus.c b/drivers/gpu/drm/i915/display/intel_gmbus.c
index 4ef8a81ae0ad..1fd3a5a6296b 100644
--- a/drivers/gpu/drm/i915/display/intel_gmbus.c
+++ b/drivers/gpu/drm/i915/display/intel_gmbus.c
@@ -32,7 +32,6 @@
 #include <linux/i2c.h>
 
 #include <drm/drm_hdcp.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_display_types.h"
@@ -632,8 +631,9 @@ retry:
 	 * till then let it sleep.
 	 */
 	if (gmbus_wait_idle(dev_priv)) {
-		DRM_DEBUG_KMS("GMBUS [%s] timed out waiting for idle\n",
-			 adapter->name);
+		drm_dbg_kms(&dev_priv->drm,
+			    "GMBUS [%s] timed out waiting for idle\n",
+			    adapter->name);
 		ret = -ETIMEDOUT;
 	}
 	intel_de_write_fw(dev_priv, GMBUS0, 0);
@@ -656,8 +656,9 @@ clear_err:
 	 */
 	ret = -ENXIO;
 	if (gmbus_wait_idle(dev_priv)) {
-		DRM_DEBUG_KMS("GMBUS [%s] timed out after NAK\n",
-			      adapter->name);
+		drm_dbg_kms(&dev_priv->drm,
+			    "GMBUS [%s] timed out after NAK\n",
+			    adapter->name);
 		ret = -ETIMEDOUT;
 	}
 
@@ -669,9 +670,9 @@ clear_err:
 	intel_de_write_fw(dev_priv, GMBUS1, 0);
 	intel_de_write_fw(dev_priv, GMBUS0, 0);
 
-	DRM_DEBUG_KMS("GMBUS [%s] NAK for addr: %04x %c(%d)\n",
-			 adapter->name, msgs[i].addr,
-			 (msgs[i].flags & I2C_M_RD) ? 'r' : 'w', msgs[i].len);
+	drm_dbg_kms(&dev_priv->drm, "GMBUS [%s] NAK for addr: %04x %c(%d)\n",
+		    adapter->name, msgs[i].addr,
+		    (msgs[i].flags & I2C_M_RD) ? 'r' : 'w', msgs[i].len);
 
 	/*
 	 * Passive adapters sometimes NAK the first probe. Retry the first
@@ -680,16 +681,18 @@ clear_err:
 	 * drm_do_probe_ddc_edid, which bails out on the first -ENXIO.
 	 */
 	if (ret == -ENXIO && i == 0 && try++ == 0) {
-		DRM_DEBUG_KMS("GMBUS [%s] NAK on first message, retry\n",
-			      adapter->name);
+		drm_dbg_kms(&dev_priv->drm,
+			    "GMBUS [%s] NAK on first message, retry\n",
+			    adapter->name);
 		goto retry;
 	}
 
 	goto out;
 
 timeout:
-	DRM_DEBUG_KMS("GMBUS [%s] timed out, falling back to bit banging on pin %d\n",
-		      bus->adapter.name, bus->reg0 & 0xff);
+	drm_dbg_kms(&dev_priv->drm,
+		    "GMBUS [%s] timed out, falling back to bit banging on pin %d\n",
+		    bus->adapter.name, bus->reg0 & 0xff);
 	intel_de_write_fw(dev_priv, GMBUS0, 0);
 
 	/*
@@ -926,9 +929,10 @@ void intel_gmbus_force_bit(struct i2c_adapter *adapter, bool force_bit)
 	mutex_lock(&dev_priv->gmbus_mutex);
 
 	bus->force_bit += force_bit ? 1 : -1;
-	DRM_DEBUG_KMS("%sabling bit-banging on %s. force bit now %d\n",
-		      force_bit ? "en" : "dis", adapter->name,
-		      bus->force_bit);
+	drm_dbg_kms(&dev_priv->drm,
+		    "%sabling bit-banging on %s. force bit now %d\n",
+		    force_bit ? "en" : "dis", adapter->name,
+		    bus->force_bit);
 
 	mutex_unlock(&dev_priv->gmbus_mutex);
 }
diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
index 229b4e329864..ee0f27ea2810 100644
--- a/drivers/gpu/drm/i915/display/intel_hdcp.c
+++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
@@ -43,6 +43,7 @@ static
 int intel_hdcp_read_valid_bksv(struct intel_digital_port *intel_dig_port,
 			       const struct intel_hdcp_shim *shim, u8 *bksv)
 {
+	struct drm_i915_private *i915 = to_i915(intel_dig_port->base.base.dev);
 	int ret, i, tries = 2;
 
 	/* HDCP spec states that we must retry the bksv if it is invalid */
@@ -54,7 +55,7 @@ int intel_hdcp_read_valid_bksv(struct intel_digital_port *intel_dig_port,
 			break;
 	}
 	if (i == tries) {
-		DRM_DEBUG_KMS("Bksv is invalid\n");
+		drm_dbg_kms(&i915->drm, "Bksv is invalid\n");
 		return -ENODEV;
 	}
 
@@ -485,8 +486,8 @@ int intel_hdcp_validate_v_prime(struct intel_connector *connector,
 			return ret;
 		sha_idx += sizeof(sha_text);
 	} else {
-		DRM_DEBUG_KMS("Invalid number of leftovers %d\n",
-			      sha_leftovers);
+		drm_dbg_kms(&dev_priv->drm, "Invalid number of leftovers %d\n",
+			    sha_leftovers);
 		return -EINVAL;
 	}
 
@@ -514,11 +515,11 @@ int intel_hdcp_validate_v_prime(struct intel_connector *connector,
 		       rep_ctl | HDCP_SHA1_COMPLETE_HASH);
 	if (intel_de_wait_for_set(dev_priv, HDCP_REP_CTL,
 				  HDCP_SHA1_COMPLETE, 1)) {
-		DRM_ERROR("Timed out waiting for SHA1 complete\n");
+		drm_err(&dev_priv->drm, "Timed out waiting for SHA1 complete\n");
 		return -ETIMEDOUT;
 	}
 	if (!(intel_de_read(dev_priv, HDCP_REP_CTL) & HDCP_SHA1_V_MATCH)) {
-		DRM_DEBUG_KMS("SHA-1 mismatch, HDCP failed\n");
+		drm_dbg_kms(&dev_priv->drm, "SHA-1 mismatch, HDCP failed\n");
 		return -ENXIO;
 	}
 
@@ -537,7 +538,8 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
 
 	ret = intel_hdcp_poll_ksv_fifo(intel_dig_port, shim);
 	if (ret) {
-		DRM_DEBUG_KMS("KSV list failed to become ready (%d)\n", ret);
+		drm_dbg_kms(&dev_priv->drm,
+			    "KSV list failed to become ready (%d)\n", ret);
 		return ret;
 	}
 
@@ -547,7 +549,7 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
 
 	if (DRM_HDCP_MAX_DEVICE_EXCEEDED(bstatus[0]) ||
 	    DRM_HDCP_MAX_CASCADE_EXCEEDED(bstatus[1])) {
-		DRM_DEBUG_KMS("Max Topology Limit Exceeded\n");
+		drm_dbg_kms(&dev_priv->drm, "Max Topology Limit Exceeded\n");
 		return -EPERM;
 	}
 
@@ -560,13 +562,14 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
 	 */
 	num_downstream = DRM_HDCP_NUM_DOWNSTREAM(bstatus[0]);
 	if (num_downstream == 0) {
-		DRM_DEBUG_KMS("Repeater with zero downstream devices\n");
+		drm_dbg_kms(&dev_priv->drm,
+			    "Repeater with zero downstream devices\n");
 		return -EINVAL;
 	}
 
 	ksv_fifo = kcalloc(DRM_HDCP_KSV_LEN, num_downstream, GFP_KERNEL);
 	if (!ksv_fifo) {
-		DRM_DEBUG_KMS("Out of mem: ksv_fifo\n");
+		drm_dbg_kms(&dev_priv->drm, "Out of mem: ksv_fifo\n");
 		return -ENOMEM;
 	}
 
@@ -576,7 +579,7 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
 
 	if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm, ksv_fifo,
 					num_downstream)) {
-		DRM_ERROR("Revoked Ksv(s) in ksv_fifo\n");
+		drm_err(&dev_priv->drm, "Revoked Ksv(s) in ksv_fifo\n");
 		ret = -EPERM;
 		goto err;
 	}
@@ -594,12 +597,13 @@ int intel_hdcp_auth_downstream(struct intel_connector *connector)
 	}
 
 	if (i == tries) {
-		DRM_DEBUG_KMS("V Prime validation failed.(%d)\n", ret);
+		drm_dbg_kms(&dev_priv->drm,
+			    "V Prime validation failed.(%d)\n", ret);
 		goto err;
 	}
 
-	DRM_DEBUG_KMS("HDCP is enabled (%d downstream devices)\n",
-		      num_downstream);
+	drm_dbg_kms(&dev_priv->drm, "HDCP is enabled (%d downstream devices)\n",
+		    num_downstream);
 	ret = 0;
 err:
 	kfree(ksv_fifo);
@@ -642,7 +646,8 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 		if (ret)
 			return ret;
 		if (!hdcp_capable) {
-			DRM_DEBUG_KMS("Panel is not HDCP capable\n");
+			drm_dbg_kms(&dev_priv->drm,
+				    "Panel is not HDCP capable\n");
 			return -EINVAL;
 		}
 	}
@@ -659,7 +664,7 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 	if (intel_de_wait_for_set(dev_priv,
 				  HDCP_STATUS(dev_priv, cpu_transcoder, port),
 				  HDCP_STATUS_AN_READY, 1)) {
-		DRM_ERROR("Timed out waiting for An\n");
+		drm_err(&dev_priv->drm, "Timed out waiting for An\n");
 		return -ETIMEDOUT;
 	}
 
@@ -680,7 +685,7 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 		return ret;
 
 	if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm, bksv.shim, 1)) {
-		DRM_ERROR("BKSV is revoked\n");
+		drm_err(&dev_priv->drm, "BKSV is revoked\n");
 		return -EPERM;
 	}
 
@@ -706,7 +711,7 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 	/* Wait for R0 ready */
 	if (wait_for(intel_de_read(dev_priv, HDCP_STATUS(dev_priv, cpu_transcoder, port)) &
 		     (HDCP_STATUS_R0_READY | HDCP_STATUS_ENC), 1)) {
-		DRM_ERROR("Timed out waiting for R0 ready\n");
+		drm_err(&dev_priv->drm, "Timed out waiting for R0 ready\n");
 		return -ETIMEDOUT;
 	}
 
@@ -743,8 +748,10 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 	}
 
 	if (i == tries) {
-		DRM_DEBUG_KMS("Timed out waiting for Ri prime match (%x)\n",
-			      intel_de_read(dev_priv, HDCP_STATUS(dev_priv, cpu_transcoder, port)));
+		drm_dbg_kms(&dev_priv->drm,
+			    "Timed out waiting for Ri prime match (%x)\n",
+			    intel_de_read(dev_priv, HDCP_STATUS(dev_priv,
+					  cpu_transcoder, port)));
 		return -ETIMEDOUT;
 	}
 
@@ -753,7 +760,7 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 				  HDCP_STATUS(dev_priv, cpu_transcoder, port),
 				  HDCP_STATUS_ENC,
 				  ENCRYPT_STATUS_CHANGE_TIMEOUT_MS)) {
-		DRM_ERROR("Timed out waiting for encryption\n");
+		drm_err(&dev_priv->drm, "Timed out waiting for encryption\n");
 		return -ETIMEDOUT;
 	}
 
@@ -765,7 +772,7 @@ static int intel_hdcp_auth(struct intel_connector *connector)
 	if (repeater_present)
 		return intel_hdcp_auth_downstream(connector);
 
-	DRM_DEBUG_KMS("HDCP is enabled (no repeater present)\n");
+	drm_dbg_kms(&dev_priv->drm, "HDCP is enabled (no repeater present)\n");
 	return 0;
 }
 
@@ -1271,7 +1278,7 @@ static int hdcp2_authentication_key_exchange(struct intel_connector *connector)
 		return ret;
 
 	if (msgs.send_cert.rx_caps[0] != HDCP_2_2_RX_CAPS_VERSION_VAL) {
-		DRM_DEBUG_KMS("cert.rx_caps dont claim HDCP2.2\n");
+		drm_dbg_kms(&dev_priv->drm, "cert.rx_caps dont claim HDCP2.2\n");
 		return -EINVAL;
 	}
 
@@ -1280,7 +1287,7 @@ static int hdcp2_authentication_key_exchange(struct intel_connector *connector)
 	if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm,
 					msgs.send_cert.cert_rx.receiver_id,
 					1)) {
-		DRM_ERROR("Receiver ID is revoked\n");
+		drm_err(&dev_priv->drm, "Receiver ID is revoked\n");
 		return -EPERM;
 	}
 
@@ -1455,7 +1462,7 @@ int hdcp2_authenticate_repeater_topology(struct intel_connector *connector)
 
 	if (HDCP_2_2_MAX_CASCADE_EXCEEDED(rx_info[1]) ||
 	    HDCP_2_2_MAX_DEVS_EXCEEDED(rx_info[1])) {
-		DRM_DEBUG_KMS("Topology Max Size Exceeded\n");
+		drm_dbg_kms(&dev_priv->drm, "Topology Max Size Exceeded\n");
 		return -EINVAL;
 	}
 
@@ -1463,9 +1470,15 @@ int hdcp2_authenticate_repeater_topology(struct intel_connector *connector)
 	seq_num_v =
 		drm_hdcp_be24_to_cpu((const u8 *)msgs.recvid_list.seq_num_v);
 
+	if (!hdcp->hdcp2_encrypted && seq_num_v) {
+		drm_dbg_kms(&dev_priv->drm,
+			    "Non zero Seq_num_v at first RecvId_List msg\n");
+		return -EINVAL;
+	}
+
 	if (seq_num_v < hdcp->seq_num_v) {
 		/* Roll over of the seq_num_v from repeater. Reauthenticate. */
-		DRM_DEBUG_KMS("Seq_num_v roll over.\n");
+		drm_dbg_kms(&dev_priv->drm, "Seq_num_v roll over.\n");
 		return -EINVAL;
 	}
 
@@ -1474,7 +1487,7 @@ int hdcp2_authenticate_repeater_topology(struct intel_connector *connector)
 	if (drm_hdcp_check_ksvs_revoked(&dev_priv->drm,
 					msgs.recvid_list.receiver_ids,
 					device_cnt)) {
-		DRM_ERROR("Revoked receiver ID(s) is in list\n");
+		drm_err(&dev_priv->drm, "Revoked receiver ID(s) is in list\n");
 		return -EPERM;
 	}
 
@@ -1507,25 +1520,27 @@ static int hdcp2_authenticate_repeater(struct intel_connector *connector)
 static int hdcp2_authenticate_sink(struct intel_connector *connector)
 {
 	struct intel_digital_port *intel_dig_port = intel_attached_dig_port(connector);
+	struct drm_i915_private *i915 = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	const struct intel_hdcp_shim *shim = hdcp->shim;
 	int ret;
 
 	ret = hdcp2_authentication_key_exchange(connector);
 	if (ret < 0) {
-		DRM_DEBUG_KMS("AKE Failed. Err : %d\n", ret);
+		drm_dbg_kms(&i915->drm, "AKE Failed. Err : %d\n", ret);
 		return ret;
 	}
 
 	ret = hdcp2_locality_check(connector);
 	if (ret < 0) {
-		DRM_DEBUG_KMS("Locality Check failed. Err : %d\n", ret);
+		drm_dbg_kms(&i915->drm,
+			    "Locality Check failed. Err : %d\n", ret);
 		return ret;
 	}
 
 	ret = hdcp2_session_key_exchange(connector);
 	if (ret < 0) {
-		DRM_DEBUG_KMS("SKE Failed. Err : %d\n", ret);
+		drm_dbg_kms(&i915->drm, "SKE Failed. Err : %d\n", ret);
 		return ret;
 	}
 
@@ -1540,7 +1555,8 @@ static int hdcp2_authenticate_sink(struct intel_connector *connector)
 	if (hdcp->is_repeater) {
 		ret = hdcp2_authenticate_repeater(connector);
 		if (ret < 0) {
-			DRM_DEBUG_KMS("Repeater Auth Failed. Err: %d\n", ret);
+			drm_dbg_kms(&i915->drm,
+				    "Repeater Auth Failed. Err: %d\n", ret);
 			return ret;
 		}
 	}
@@ -1630,6 +1646,7 @@ static int hdcp2_disable_encryption(struct intel_connector *connector)
 
 static int hdcp2_authenticate_and_encrypt(struct intel_connector *connector)
 {
+	struct drm_i915_private *i915 = to_i915(connector->base.dev);
 	int ret, i, tries = 3;
 
 	for (i = 0; i < tries; i++) {
@@ -1638,10 +1655,10 @@ static int hdcp2_authenticate_and_encrypt(struct intel_connector *connector)
 			break;
 
 		/* Clearing the mei hdcp session */
-		DRM_DEBUG_KMS("HDCP2.2 Auth %d of %d Failed.(%d)\n",
-			      i + 1, tries, ret);
+		drm_dbg_kms(&i915->drm, "HDCP2.2 Auth %d of %d Failed.(%d)\n",
+			    i + 1, tries, ret);
 		if (hdcp2_deauthenticate_port(connector) < 0)
-			DRM_DEBUG_KMS("Port deauth failed.\n");
+			drm_dbg_kms(&i915->drm, "Port deauth failed.\n");
 	}
 
 	if (i != tries) {
@@ -1652,9 +1669,10 @@ static int hdcp2_authenticate_and_encrypt(struct intel_connector *connector)
 		msleep(HDCP_2_2_DELAY_BEFORE_ENCRYPTION_EN);
 		ret = hdcp2_enable_encryption(connector);
 		if (ret < 0) {
-			DRM_DEBUG_KMS("Encryption Enable Failed.(%d)\n", ret);
+			drm_dbg_kms(&i915->drm,
+				    "Encryption Enable Failed.(%d)\n", ret);
 			if (hdcp2_deauthenticate_port(connector) < 0)
-				DRM_DEBUG_KMS("Port deauth failed.\n");
+				drm_dbg_kms(&i915->drm, "Port deauth failed.\n");
 		}
 	}
 
@@ -1663,23 +1681,24 @@ static int hdcp2_authenticate_and_encrypt(struct intel_connector *connector)
 
 static int _intel_hdcp2_enable(struct intel_connector *connector)
 {
+	struct drm_i915_private *i915 = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	int ret;
 
-	DRM_DEBUG_KMS("[%s:%d] HDCP2.2 is being enabled. Type: %d\n",
-		      connector->base.name, connector->base.base.id,
-		      hdcp->content_type);
+	drm_dbg_kms(&i915->drm, "[%s:%d] HDCP2.2 is being enabled. Type: %d\n",
+		    connector->base.name, connector->base.base.id,
+		    hdcp->content_type);
 
 	ret = hdcp2_authenticate_and_encrypt(connector);
 	if (ret) {
-		DRM_DEBUG_KMS("HDCP2 Type%d  Enabling Failed. (%d)\n",
-			      hdcp->content_type, ret);
+		drm_dbg_kms(&i915->drm, "HDCP2 Type%d  Enabling Failed. (%d)\n",
+			    hdcp->content_type, ret);
 		return ret;
 	}
 
-	DRM_DEBUG_KMS("[%s:%d] HDCP2.2 is enabled. Type %d\n",
-		      connector->base.name, connector->base.base.id,
-		      hdcp->content_type);
+	drm_dbg_kms(&i915->drm, "[%s:%d] HDCP2.2 is enabled. Type %d\n",
+		    connector->base.name, connector->base.base.id,
+		    hdcp->content_type);
 
 	hdcp->hdcp2_encrypted = true;
 	return 0;
@@ -1687,15 +1706,16 @@ static int _intel_hdcp2_enable(struct intel_connector *connector)
 
 static int _intel_hdcp2_disable(struct intel_connector *connector)
 {
+	struct drm_i915_private *i915 = to_i915(connector->base.dev);
 	int ret;
 
-	DRM_DEBUG_KMS("[%s:%d] HDCP2.2 is being Disabled\n",
-		      connector->base.name, connector->base.base.id);
+	drm_dbg_kms(&i915->drm, "[%s:%d] HDCP2.2 is being Disabled\n",
+		    connector->base.name, connector->base.base.id);
 
 	ret = hdcp2_disable_encryption(connector);
 
 	if (hdcp2_deauthenticate_port(connector) < 0)
-		DRM_DEBUG_KMS("Port deauth failed.\n");
+		drm_dbg_kms(&i915->drm, "Port deauth failed.\n");
 
 	connector->hdcp.hdcp2_encrypted = false;
 
@@ -1938,12 +1958,13 @@ void intel_hdcp_component_init(struct drm_i915_private *dev_priv)
 static void intel_hdcp2_init(struct intel_connector *connector,
 			     const struct intel_hdcp_shim *shim)
 {
+	struct drm_i915_private *i915 = to_i915(connector->base.dev);
 	struct intel_hdcp *hdcp = &connector->hdcp;
 	int ret;
 
 	ret = initialize_hdcp_port_data(connector, shim);
 	if (ret) {
-		DRM_DEBUG_KMS("Mei hdcp data init failed\n");
+		drm_dbg_kms(&i915->drm, "Mei hdcp data init failed\n");
 		return;
 	}
 
diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.h b/drivers/gpu/drm/i915/display/intel_hdcp.h
index 1b2eacaf8949..7c12ad609b1f 100644
--- a/drivers/gpu/drm/i915/display/intel_hdcp.h
+++ b/drivers/gpu/drm/i915/display/intel_hdcp.h
@@ -8,8 +8,6 @@
 
 #include <linux/types.h>
 
-#include <drm/i915_drm.h>
-
 struct drm_connector;
 struct drm_connector_state;
 struct drm_i915_private;
diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.c b/drivers/gpu/drm/i915/display/intel_hdmi.c
index bdbb5ce3fa81..39930232b253 100644
--- a/drivers/gpu/drm/i915/display/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/display/intel_hdmi.c
@@ -36,7 +36,6 @@
 #include <drm/drm_edid.h>
 #include <drm/drm_hdcp.h>
 #include <drm/drm_scdc_helper.h>
-#include <drm/i915_drm.h>
 #include <drm/intel_lpe_audio.h>
 
 #include "i915_debugfs.h"
@@ -2276,14 +2275,9 @@ static bool hdmi_deep_color_possible(const struct intel_crtc_state *crtc_state,
 		}
 	}
 
-	/* Display WA #1139: glk */
-	if (bpc == 12 && IS_GLK_REVID(dev_priv, 0, GLK_REVID_A1) &&
-	    adjusted_mode->htotal > 5460)
-		return false;
-
-	/* Display Wa_1405510057:icl */
+	/* Display Wa_1405510057:icl,ehl */
 	if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420 &&
-	    bpc == 10 && INTEL_GEN(dev_priv) >= 11 &&
+	    bpc == 10 && IS_GEN(dev_priv, 11) &&
 	    (adjusted_mode->crtc_hblank_end -
 	     adjusted_mode->crtc_hblank_start) % 8 == 2)
 		return false;
diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.h b/drivers/gpu/drm/i915/display/intel_hdmi.h
index d3659d0b408b..8ff1f76a63df 100644
--- a/drivers/gpu/drm/i915/display/intel_hdmi.h
+++ b/drivers/gpu/drm/i915/display/intel_hdmi.h
@@ -9,8 +9,6 @@
 #include <linux/hdmi.h>
 #include <linux/types.h>
 
-#include <drm/i915_drm.h>
-
 #include "i915_reg.h"
 
 struct drm_connector;
diff --git a/drivers/gpu/drm/i915/display/intel_hotplug.c b/drivers/gpu/drm/i915/display/intel_hotplug.c
index 8af0ae61e1bb..a091442efba4 100644
--- a/drivers/gpu/drm/i915/display/intel_hotplug.c
+++ b/drivers/gpu/drm/i915/display/intel_hotplug.c
@@ -23,8 +23,6 @@
 
 #include <linux/kernel.h>
 
-#include <drm/i915_drm.h>
-
 #include "i915_drv.h"
 #include "intel_display_types.h"
 #include "intel_hotplug.h"
@@ -89,29 +87,16 @@
 enum hpd_pin intel_hpd_pin_default(struct drm_i915_private *dev_priv,
 				   enum port port)
 {
-	switch (port) {
-	case PORT_A:
-		return HPD_PORT_A;
-	case PORT_B:
-		return HPD_PORT_B;
-	case PORT_C:
-		return HPD_PORT_C;
-	case PORT_D:
-		return HPD_PORT_D;
-	case PORT_E:
-		return HPD_PORT_E;
-	case PORT_F:
-		if (IS_CNL_WITH_PORT_F(dev_priv))
-			return HPD_PORT_E;
-		return HPD_PORT_F;
-	case PORT_G:
-		return HPD_PORT_G;
-	case PORT_H:
-		return HPD_PORT_H;
-	case PORT_I:
-		return HPD_PORT_I;
+	enum phy phy = intel_port_to_phy(dev_priv, port);
+
+	switch (phy) {
+	case PHY_F:
+		return IS_CNL_WITH_PORT_F(dev_priv) ? HPD_PORT_E : HPD_PORT_F;
+	case PHY_A ... PHY_E:
+	case PHY_G ... PHY_I:
+		return HPD_PORT_A + phy - PHY_A;
 	default:
-		MISSING_CASE(port);
+		MISSING_CASE(phy);
 		return HPD_NONE;
 	}
 }
@@ -185,10 +170,13 @@ static bool intel_hpd_irq_storm_detect(struct drm_i915_private *dev_priv,
 	hpd->stats[pin].count += increment;
 	if (hpd->stats[pin].count > threshold) {
 		hpd->stats[pin].state = HPD_MARK_DISABLED;
-		DRM_DEBUG_KMS("HPD interrupt storm detected on PIN %d\n", pin);
+		drm_dbg_kms(&dev_priv->drm,
+			    "HPD interrupt storm detected on PIN %d\n", pin);
 		storm = true;
 	} else {
-		DRM_DEBUG_KMS("Received HPD interrupt on PIN %d - cnt: %d\n", pin,
+		drm_dbg_kms(&dev_priv->drm,
+			    "Received HPD interrupt on PIN %d - cnt: %d\n",
+			      pin,
 			      hpd->stats[pin].count);
 	}
 
@@ -217,7 +205,8 @@ intel_hpd_irq_storm_switch_to_polling(struct drm_i915_private *dev_priv)
 		    dev_priv->hotplug.stats[pin].state != HPD_MARK_DISABLED)
 			continue;
 
-		DRM_INFO("HPD interrupt storm detected on connector %s: "
+		drm_info(&dev_priv->drm,
+			 "HPD interrupt storm detected on connector %s: "
 			 "switching from hotplug detection to polling\n",
 			 connector->base.name);
 
@@ -259,8 +248,9 @@ static void intel_hpd_irq_storm_reenable_work(struct work_struct *work)
 			continue;
 
 		if (connector->base.polled != connector->polled)
-			DRM_DEBUG_DRIVER("Reenabling HPD on connector %s\n",
-					 connector->base.name);
+			drm_dbg(&dev_priv->drm,
+				"Reenabling HPD on connector %s\n",
+				connector->base.name);
 		connector->base.polled = connector->polled;
 	}
 	drm_connector_list_iter_end(&conn_iter);
@@ -295,11 +285,12 @@ intel_encoder_hotplug(struct intel_encoder *encoder,
 	if (old_status == connector->base.status)
 		return INTEL_HOTPLUG_UNCHANGED;
 
-	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n",
-		      connector->base.base.id,
-		      connector->base.name,
-		      drm_get_connector_status_name(old_status),
-		      drm_get_connector_status_name(connector->base.status));
+	drm_dbg_kms(&to_i915(dev)->drm,
+		    "[CONNECTOR:%d:%s] status updated from %s to %s\n",
+		    connector->base.base.id,
+		    connector->base.name,
+		    drm_get_connector_status_name(old_status),
+		    drm_get_connector_status_name(connector->base.status));
 
 	return INTEL_HOTPLUG_CHANGED;
 }
@@ -373,7 +364,7 @@ static void i915_hotplug_work_func(struct work_struct *work)
 	u32 hpd_retry_bits;
 
 	mutex_lock(&dev->mode_config.mutex);
-	DRM_DEBUG_KMS("running encoder hotplug functions\n");
+	drm_dbg_kms(&dev_priv->drm, "running encoder hotplug functions\n");
 
 	spin_lock_irq(&dev_priv->irq_lock);
 
@@ -401,8 +392,9 @@ static void i915_hotplug_work_func(struct work_struct *work)
 			struct intel_encoder *encoder =
 				intel_attached_encoder(connector);
 
-			DRM_DEBUG_KMS("Connector %s (pin %i) received hotplug event.\n",
-				      connector->base.name, pin);
+			drm_dbg_kms(&dev_priv->drm,
+				    "Connector %s (pin %i) received hotplug event.\n",
+				    connector->base.name, pin);
 
 			switch (encoder->hotplug(encoder, connector,
 						 hpd_event_bits & hpd_bit)) {
@@ -487,9 +479,10 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv,
 
 		long_hpd = long_mask & BIT(pin);
 
-		DRM_DEBUG_DRIVER("digital hpd on [ENCODER:%d:%s] - %s\n",
-				 encoder->base.base.id, encoder->base.name,
-				 long_hpd ? "long" : "short");
+		drm_dbg(&dev_priv->drm,
+			"digital hpd on [ENCODER:%d:%s] - %s\n",
+			encoder->base.base.id, encoder->base.name,
+			long_hpd ? "long" : "short");
 		queue_dig = true;
 
 		if (long_hpd) {
diff --git a/drivers/gpu/drm/i915/display/intel_hotplug.h b/drivers/gpu/drm/i915/display/intel_hotplug.h
index 087b5f57b321..1e6b4fda2900 100644
--- a/drivers/gpu/drm/i915/display/intel_hotplug.h
+++ b/drivers/gpu/drm/i915/display/intel_hotplug.h
@@ -8,8 +8,6 @@
 
 #include <linux/types.h>
 
-#include <drm/i915_drm.h>
-
 struct drm_i915_private;
 struct intel_connector;
 struct intel_encoder;
diff --git a/drivers/gpu/drm/i915/display/intel_lpe_audio.c b/drivers/gpu/drm/i915/display/intel_lpe_audio.c
index 516e7179a5a4..ad5cc13037ae 100644
--- a/drivers/gpu/drm/i915/display/intel_lpe_audio.c
+++ b/drivers/gpu/drm/i915/display/intel_lpe_audio.c
@@ -127,7 +127,8 @@ lpe_audio_platdev_create(struct drm_i915_private *dev_priv)
 	kfree(pdata);
 
 	if (IS_ERR(platdev)) {
-		DRM_ERROR("Failed to allocate LPE audio platform device\n");
+		drm_err(&dev_priv->drm,
+			"Failed to allocate LPE audio platform device\n");
 		return platdev;
 	}
 
@@ -190,7 +191,8 @@ static bool lpe_audio_detect(struct drm_i915_private *dev_priv)
 		};
 
 		if (!pci_dev_present(atom_hdaudio_ids)) {
-			DRM_INFO("HDaudio controller not detected, using LPE audio instead\n");
+			drm_info(&dev_priv->drm,
+				 "HDaudio controller not detected, using LPE audio instead\n");
 			lpe_present = true;
 		}
 	}
@@ -203,18 +205,19 @@ static int lpe_audio_setup(struct drm_i915_private *dev_priv)
 
 	dev_priv->lpe_audio.irq = irq_alloc_desc(0);
 	if (dev_priv->lpe_audio.irq < 0) {
-		DRM_ERROR("Failed to allocate IRQ desc: %d\n",
+		drm_err(&dev_priv->drm, "Failed to allocate IRQ desc: %d\n",
 			dev_priv->lpe_audio.irq);
 		ret = dev_priv->lpe_audio.irq;
 		goto err;
 	}
 
-	DRM_DEBUG("irq = %d\n", dev_priv->lpe_audio.irq);
+	drm_dbg(&dev_priv->drm, "irq = %d\n", dev_priv->lpe_audio.irq);
 
 	ret = lpe_audio_irq_init(dev_priv);
 
 	if (ret) {
-		DRM_ERROR("Failed to initialize irqchip for lpe audio: %d\n",
+		drm_err(&dev_priv->drm,
+			"Failed to initialize irqchip for lpe audio: %d\n",
 			ret);
 		goto err_free_irq;
 	}
@@ -223,7 +226,8 @@ static int lpe_audio_setup(struct drm_i915_private *dev_priv)
 
 	if (IS_ERR(dev_priv->lpe_audio.platdev)) {
 		ret = PTR_ERR(dev_priv->lpe_audio.platdev);
-		DRM_ERROR("Failed to create lpe audio platform device: %d\n",
+		drm_err(&dev_priv->drm,
+			"Failed to create lpe audio platform device: %d\n",
 			ret);
 		goto err_free_irq;
 	}
@@ -259,8 +263,8 @@ void intel_lpe_audio_irq_handler(struct drm_i915_private *dev_priv)
 
 	ret = generic_handle_irq(dev_priv->lpe_audio.irq);
 	if (ret)
-		DRM_ERROR_RATELIMITED("error handling LPE audio irq: %d\n",
-				ret);
+		drm_err_ratelimited(&dev_priv->drm,
+				    "error handling LPE audio irq: %d\n", ret);
 }
 
 /**
@@ -278,7 +282,8 @@ int intel_lpe_audio_init(struct drm_i915_private *dev_priv)
 	if (lpe_audio_detect(dev_priv)) {
 		ret = lpe_audio_setup(dev_priv);
 		if (ret < 0)
-			DRM_ERROR("failed to setup LPE Audio bridge\n");
+			drm_err(&dev_priv->drm,
+				"failed to setup LPE Audio bridge\n");
 	}
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/display/intel_lvds.c b/drivers/gpu/drm/i915/display/intel_lvds.c
index b7ad0b534790..9a067effcfa0 100644
--- a/drivers/gpu/drm/i915/display/intel_lvds.c
+++ b/drivers/gpu/drm/i915/display/intel_lvds.c
@@ -37,7 +37,6 @@
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_atomic.h"
@@ -182,8 +181,9 @@ static void intel_lvds_pps_get_hw_state(struct drm_i915_private *dev_priv,
 
 	if (INTEL_GEN(dev_priv) <= 4 &&
 	    pps->t1_t2 == 0 && pps->t5 == 0 && pps->t3 == 0 && pps->tx == 0) {
-		DRM_DEBUG_KMS("Panel power timings uninitialized, "
-			      "setting defaults\n");
+		drm_dbg_kms(&dev_priv->drm,
+			    "Panel power timings uninitialized, "
+			    "setting defaults\n");
 		/* Set T2 to 40ms and T5 to 200ms in 100 usec units */
 		pps->t1_t2 = 40 * 10;
 		pps->t5 = 200 * 10;
@@ -192,10 +192,10 @@ static void intel_lvds_pps_get_hw_state(struct drm_i915_private *dev_priv,
 		pps->tx = 200 * 10;
 	}
 
-	DRM_DEBUG_DRIVER("LVDS PPS:t1+t2 %d t3 %d t4 %d t5 %d tx %d "
-			 "divider %d port %d powerdown_on_reset %d\n",
-			 pps->t1_t2, pps->t3, pps->t4, pps->t5, pps->tx,
-			 pps->divider, pps->port, pps->powerdown_on_reset);
+	drm_dbg(&dev_priv->drm, "LVDS PPS:t1+t2 %d t3 %d t4 %d t5 %d tx %d "
+		"divider %d port %d powerdown_on_reset %d\n",
+		pps->t1_t2, pps->t3, pps->t4, pps->t5, pps->tx,
+		pps->divider, pps->port, pps->powerdown_on_reset);
 }
 
 static void intel_lvds_pps_init_hw(struct drm_i915_private *dev_priv,
@@ -317,7 +317,8 @@ static void intel_enable_lvds(struct intel_encoder *encoder,
 	intel_de_posting_read(dev_priv, lvds_encoder->reg);
 
 	if (intel_de_wait_for_set(dev_priv, PP_STATUS(0), PP_ON, 5000))
-		DRM_ERROR("timed out waiting for panel to power on\n");
+		drm_err(&dev_priv->drm,
+			"timed out waiting for panel to power on\n");
 
 	intel_panel_enable_backlight(pipe_config, conn_state);
 }
@@ -332,7 +333,8 @@ static void intel_disable_lvds(struct intel_encoder *encoder,
 	intel_de_write(dev_priv, PP_CONTROL(0),
 		       intel_de_read(dev_priv, PP_CONTROL(0)) & ~PANEL_POWER_ON);
 	if (intel_de_wait_for_clear(dev_priv, PP_STATUS(0), PP_ON, 1000))
-		DRM_ERROR("timed out waiting for panel to power off\n");
+		drm_err(&dev_priv->drm,
+			"timed out waiting for panel to power off\n");
 
 	intel_de_write(dev_priv, lvds_encoder->reg,
 		       intel_de_read(dev_priv, lvds_encoder->reg) & ~LVDS_PORT_EN);
@@ -398,7 +400,7 @@ static int intel_lvds_compute_config(struct intel_encoder *intel_encoder,
 
 	/* Should never happen!! */
 	if (INTEL_GEN(dev_priv) < 4 && intel_crtc->pipe == 0) {
-		DRM_ERROR("Can't support LVDS on pipe A\n");
+		drm_err(&dev_priv->drm, "Can't support LVDS on pipe A\n");
 		return -EINVAL;
 	}
 
@@ -408,8 +410,9 @@ static int intel_lvds_compute_config(struct intel_encoder *intel_encoder,
 		lvds_bpp = 6*3;
 
 	if (lvds_bpp != pipe_config->pipe_bpp && !pipe_config->bw_constrained) {
-		DRM_DEBUG_KMS("forcing display bpp (was %d) to LVDS (%d)\n",
-			      pipe_config->pipe_bpp, lvds_bpp);
+		drm_dbg_kms(&dev_priv->drm,
+			    "forcing display bpp (was %d) to LVDS (%d)\n",
+			    pipe_config->pipe_bpp, lvds_bpp);
 		pipe_config->pipe_bpp = lvds_bpp;
 	}
 
@@ -833,7 +836,8 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 	}
 
 	if (!dev_priv->vbt.int_lvds_support) {
-		DRM_DEBUG_KMS("Internal LVDS support disabled by VBT\n");
+		drm_dbg_kms(&dev_priv->drm,
+			    "Internal LVDS support disabled by VBT\n");
 		return;
 	}
 
@@ -852,10 +856,12 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 	pin = GMBUS_PIN_PANEL;
 	if (!intel_bios_is_lvds_present(dev_priv, &pin)) {
 		if ((lvds & LVDS_PORT_EN) == 0) {
-			DRM_DEBUG_KMS("LVDS is not present in VBT\n");
+			drm_dbg_kms(&dev_priv->drm,
+				    "LVDS is not present in VBT\n");
 			return;
 		}
-		DRM_DEBUG_KMS("LVDS is not present in VBT, but enabled anyway\n");
+		drm_dbg_kms(&dev_priv->drm,
+			    "LVDS is not present in VBT, but enabled anyway\n");
 	}
 
 	lvds_encoder = kzalloc(sizeof(*lvds_encoder), GFP_KERNEL);
@@ -969,7 +975,7 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 	 */
 	fixed_mode = intel_encoder_current_mode(intel_encoder);
 	if (fixed_mode) {
-		DRM_DEBUG_KMS("using current (BIOS) mode: ");
+		drm_dbg_kms(&dev_priv->drm, "using current (BIOS) mode: ");
 		drm_mode_debug_printmodeline(fixed_mode);
 		fixed_mode->type |= DRM_MODE_TYPE_PREFERRED;
 	}
@@ -985,8 +991,8 @@ out:
 	intel_panel_setup_backlight(connector, INVALID_PIPE);
 
 	lvds_encoder->is_dual_link = compute_is_dual_link_lvds(lvds_encoder);
-	DRM_DEBUG_KMS("detected %s-link lvds configuration\n",
-		      lvds_encoder->is_dual_link ? "dual" : "single");
+	drm_dbg_kms(&dev_priv->drm, "detected %s-link lvds configuration\n",
+		    lvds_encoder->is_dual_link ? "dual" : "single");
 
 	lvds_encoder->a3_power = lvds & LVDS_A3_POWER_MASK;
 
@@ -995,7 +1001,7 @@ out:
 failed:
 	mutex_unlock(&dev->mode_config.mutex);
 
-	DRM_DEBUG_KMS("No LVDS modes found, disabling.\n");
+	drm_dbg_kms(&dev_priv->drm, "No LVDS modes found, disabling.\n");
 	drm_connector_cleanup(connector);
 	drm_encoder_cleanup(encoder);
 	kfree(lvds_encoder);
diff --git a/drivers/gpu/drm/i915/display/intel_opregion.c b/drivers/gpu/drm/i915/display/intel_opregion.c
index dfd78fccd456..cc6b00959586 100644
--- a/drivers/gpu/drm/i915/display/intel_opregion.c
+++ b/drivers/gpu/drm/i915/display/intel_opregion.c
@@ -30,8 +30,6 @@
 #include <linux/firmware.h>
 #include <acpi/video.h>
 
-#include <drm/i915_drm.h>
-
 #include "display/intel_panel.h"
 
 #include "i915_drv.h"
diff --git a/drivers/gpu/drm/i915/display/intel_overlay.c b/drivers/gpu/drm/i915/display/intel_overlay.c
index 5f1207dec10e..481187223101 100644
--- a/drivers/gpu/drm/i915/display/intel_overlay.c
+++ b/drivers/gpu/drm/i915/display/intel_overlay.c
@@ -27,7 +27,6 @@
  */
 
 #include <drm/drm_fourcc.h>
-#include <drm/i915_drm.h>
 
 #include "gem/i915_gem_pm.h"
 #include "gt/intel_ring.h"
@@ -324,7 +323,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 	/* check for underruns */
 	tmp = intel_de_read(dev_priv, DOVSTA);
 	if (tmp & (1 << 17))
-		DRM_DEBUG("overlay underrun, DOVSTA: %x\n", tmp);
+		drm_dbg(&dev_priv->drm, "overlay underrun, DOVSTA: %x\n", tmp);
 
 	rq = alloc_request(overlay, NULL);
 	if (IS_ERR(rq))
@@ -1069,7 +1068,7 @@ int intel_overlay_put_image_ioctl(struct drm_device *dev, void *data,
 
 	overlay = dev_priv->overlay;
 	if (!overlay) {
-		DRM_DEBUG("userspace bug: no overlay\n");
+		drm_dbg(&dev_priv->drm, "userspace bug: no overlay\n");
 		return -ENODEV;
 	}
 
@@ -1093,7 +1092,8 @@ int intel_overlay_put_image_ioctl(struct drm_device *dev, void *data,
 	drm_modeset_lock_all(dev);
 
 	if (i915_gem_object_is_tiled(new_bo)) {
-		DRM_DEBUG_KMS("buffer used for overlay image can not be tiled\n");
+		drm_dbg_kms(&dev_priv->drm,
+			    "buffer used for overlay image can not be tiled\n");
 		ret = -EINVAL;
 		goto out_unlock;
 	}
@@ -1228,7 +1228,7 @@ int intel_overlay_attrs_ioctl(struct drm_device *dev, void *data,
 
 	overlay = dev_priv->overlay;
 	if (!overlay) {
-		DRM_DEBUG("userspace bug: no overlay\n");
+		drm_dbg(&dev_priv->drm, "userspace bug: no overlay\n");
 		return -ENODEV;
 	}
 
@@ -1372,7 +1372,7 @@ void intel_overlay_setup(struct drm_i915_private *dev_priv)
 	update_reg_attrs(overlay, overlay->regs);
 
 	dev_priv->overlay = overlay;
-	DRM_INFO("Initialized overlay support.\n");
+	drm_info(&dev_priv->drm, "Initialized overlay support.\n");
 	return;
 
 out_free:
diff --git a/drivers/gpu/drm/i915/display/intel_panel.c b/drivers/gpu/drm/i915/display/intel_panel.c
index 5aead622019c..276f43870802 100644
--- a/drivers/gpu/drm/i915/display/intel_panel.c
+++ b/drivers/gpu/drm/i915/display/intel_panel.c
@@ -1882,6 +1882,7 @@ static int pwm_setup_backlight(struct intel_connector *connector,
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_panel *panel = &connector->panel;
 	const char *desc;
+	u32 level, ns;
 	int retval;
 
 	/* Get the right PWM chip for DSI backlight according to VBT */
@@ -1906,8 +1907,12 @@ static int pwm_setup_backlight(struct intel_connector *connector,
 	 */
 	pwm_apply_args(panel->backlight.pwm);
 
-	retval = pwm_config(panel->backlight.pwm, CRC_PMIC_PWM_PERIOD_NS,
-			    CRC_PMIC_PWM_PERIOD_NS);
+	panel->backlight.min = 0; /* 0% */
+	panel->backlight.max = 100; /* 100% */
+	level = intel_panel_compute_brightness(connector, 100);
+	ns = DIV_ROUND_UP(level * CRC_PMIC_PWM_PERIOD_NS, 100);
+
+	retval = pwm_config(panel->backlight.pwm, ns, CRC_PMIC_PWM_PERIOD_NS);
 	if (retval < 0) {
 		drm_err(&dev_priv->drm, "Failed to configure the pwm chip\n");
 		pwm_put(panel->backlight.pwm);
@@ -1915,11 +1920,10 @@ static int pwm_setup_backlight(struct intel_connector *connector,
 		return retval;
 	}
 
-	panel->backlight.min = 0; /* 0% */
-	panel->backlight.max = 100; /* 100% */
-	panel->backlight.level = DIV_ROUND_UP(
-				 pwm_get_duty_cycle(panel->backlight.pwm) * 100,
-				 CRC_PMIC_PWM_PERIOD_NS);
+	level = DIV_ROUND_UP(pwm_get_duty_cycle(panel->backlight.pwm) * 100,
+			     CRC_PMIC_PWM_PERIOD_NS);
+	panel->backlight.level =
+		intel_panel_compute_brightness(connector, level);
 	panel->backlight.enabled = panel->backlight.level != 0;
 
 	drm_info(&dev_priv->drm, "Using %s PWM for LCD backlight control\n",
diff --git a/drivers/gpu/drm/i915/display/intel_pipe_crc.c b/drivers/gpu/drm/i915/display/intel_pipe_crc.c
index 59d7e3cb3445..a9a5df2fee4d 100644
--- a/drivers/gpu/drm/i915/display/intel_pipe_crc.c
+++ b/drivers/gpu/drm/i915/display/intel_pipe_crc.c
@@ -441,15 +441,11 @@ display_crc_ctl_parse_source(const char *buf, enum intel_pipe_crc_source *s)
 	return 0;
 }
 
-void intel_display_crc_init(struct drm_i915_private *dev_priv)
+void intel_crtc_crc_init(struct intel_crtc *crtc)
 {
-	enum pipe pipe;
+	struct intel_pipe_crc *pipe_crc = &crtc->pipe_crc;
 
-	for_each_pipe(dev_priv, pipe) {
-		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
-
-		spin_lock_init(&pipe_crc->lock);
-	}
+	spin_lock_init(&pipe_crc->lock);
 }
 
 static int i8xx_crc_source_valid(struct drm_i915_private *dev_priv,
@@ -587,7 +583,8 @@ int intel_crtc_verify_crc_source(struct drm_crtc *crtc, const char *source_name,
 int intel_crtc_set_crc_source(struct drm_crtc *crtc, const char *source_name)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[crtc->index];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_pipe_crc *pipe_crc = &intel_crtc->pipe_crc;
 	enum intel_display_power_domain power_domain;
 	enum intel_pipe_crc_source source;
 	intel_wakeref_t wakeref;
@@ -640,7 +637,7 @@ void intel_crtc_enable_pipe_crc(struct intel_crtc *intel_crtc)
 {
 	struct drm_crtc *crtc = &intel_crtc->base;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[crtc->index];
+	struct intel_pipe_crc *pipe_crc = &intel_crtc->pipe_crc;
 	u32 val = 0;
 
 	if (!crtc->crc.opened)
@@ -660,7 +657,7 @@ void intel_crtc_disable_pipe_crc(struct intel_crtc *intel_crtc)
 {
 	struct drm_crtc *crtc = &intel_crtc->base;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[crtc->index];
+	struct intel_pipe_crc *pipe_crc = &intel_crtc->pipe_crc;
 
 	/* Swallow crc's until we stop generating them. */
 	spin_lock_irq(&pipe_crc->lock);
diff --git a/drivers/gpu/drm/i915/display/intel_pipe_crc.h b/drivers/gpu/drm/i915/display/intel_pipe_crc.h
index db258a756fc6..43012b189415 100644
--- a/drivers/gpu/drm/i915/display/intel_pipe_crc.h
+++ b/drivers/gpu/drm/i915/display/intel_pipe_crc.h
@@ -13,7 +13,7 @@ struct drm_i915_private;
 struct intel_crtc;
 
 #ifdef CONFIG_DEBUG_FS
-void intel_display_crc_init(struct drm_i915_private *dev_priv);
+void intel_crtc_crc_init(struct intel_crtc *crtc);
 int intel_crtc_set_crc_source(struct drm_crtc *crtc, const char *source_name);
 int intel_crtc_verify_crc_source(struct drm_crtc *crtc,
 				 const char *source_name, size_t *values_cnt);
@@ -22,7 +22,7 @@ const char *const *intel_crtc_get_crc_sources(struct drm_crtc *crtc,
 void intel_crtc_disable_pipe_crc(struct intel_crtc *crtc);
 void intel_crtc_enable_pipe_crc(struct intel_crtc *crtc);
 #else
-static inline void intel_display_crc_init(struct drm_i915_private *dev_priv) {}
+static inline void intel_crtc_crc_init(struct intel_crtc *crtc) {}
 #define intel_crtc_set_crc_source NULL
 #define intel_crtc_verify_crc_source NULL
 #define intel_crtc_get_crc_sources NULL
diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c
index 7e754201f54d..fd9b146e3aba 100644
--- a/drivers/gpu/drm/i915/display/intel_psr.c
+++ b/drivers/gpu/drm/i915/display/intel_psr.c
@@ -304,7 +304,7 @@ void intel_psr_init_dpcd(struct intel_dp *intel_dp)
 	drm_dbg_kms(&dev_priv->drm, "eDP panel supports PSR version %x\n",
 		    intel_dp->psr_dpcd[0]);
 
-	if (drm_dp_has_quirk(&intel_dp->desc, DP_DPCD_QUIRK_NO_PSR)) {
+	if (drm_dp_has_quirk(&intel_dp->desc, 0, DP_DPCD_QUIRK_NO_PSR)) {
 		drm_dbg_kms(&dev_priv->drm,
 			    "PSR support not currently available for this panel\n");
 		return;
diff --git a/drivers/gpu/drm/i915/display/intel_quirks.c b/drivers/gpu/drm/i915/display/intel_quirks.c
index 9d235d270dac..46beb155d835 100644
--- a/drivers/gpu/drm/i915/display/intel_quirks.c
+++ b/drivers/gpu/drm/i915/display/intel_quirks.c
@@ -82,6 +82,16 @@ static const struct intel_dmi_quirk intel_dmi_quirks[] = {
 					    DMI_MATCH(DMI_PRODUCT_NAME, ""),
 				},
 			},
+			{
+				.callback = intel_dmi_reverse_brightness,
+				.ident = "Thundersoft TST178 tablet",
+				/* DMI strings are too generic, also match on BIOS date */
+				.matches = {DMI_EXACT_MATCH(DMI_BOARD_VENDOR, "AMI Corporation"),
+					    DMI_EXACT_MATCH(DMI_BOARD_NAME, "Aptio CRB"),
+					    DMI_EXACT_MATCH(DMI_PRODUCT_NAME, "To be filled by O.E.M."),
+					    DMI_EXACT_MATCH(DMI_BIOS_DATE, "04/15/2014"),
+				},
+			},
 			{ }  /* terminating entry */
 		},
 		.hook = quirk_invert_brightness,
diff --git a/drivers/gpu/drm/i915/display/intel_sdvo.c b/drivers/gpu/drm/i915/display/intel_sdvo.c
index b0588150752c..637d8fe2f8c2 100644
--- a/drivers/gpu/drm/i915/display/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/display/intel_sdvo.c
@@ -34,7 +34,6 @@
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_atomic.h"
diff --git a/drivers/gpu/drm/i915/display/intel_sdvo.h b/drivers/gpu/drm/i915/display/intel_sdvo.h
index a66f224aa17d..72065e4360d5 100644
--- a/drivers/gpu/drm/i915/display/intel_sdvo.h
+++ b/drivers/gpu/drm/i915/display/intel_sdvo.h
@@ -8,8 +8,6 @@
 
 #include <linux/types.h>
 
-#include <drm/i915_drm.h>
-
 #include "i915_reg.h"
 
 struct drm_i915_private;
diff --git a/drivers/gpu/drm/i915/display/intel_sprite.c b/drivers/gpu/drm/i915/display/intel_sprite.c
index 7abeefe8dce5..deda351719db 100644
--- a/drivers/gpu/drm/i915/display/intel_sprite.c
+++ b/drivers/gpu/drm/i915/display/intel_sprite.c
@@ -37,10 +37,10 @@
 #include <drm/drm_fourcc.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_rect.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "i915_trace.h"
+#include "i915_vgpu.h"
 #include "intel_atomic_plane.h"
 #include "intel_display_types.h"
 #include "intel_frontbuffer.h"
@@ -284,6 +284,16 @@ int intel_plane_check_src_coordinates(struct intel_plane_state *plane_state)
 	bool rotated = drm_rotation_90_or_270(plane_state->hw.rotation);
 
 	/*
+	 * FIXME hsub/vsub vs. block size is a mess. Pre-tgl CCS
+	 * abuses hsub/vsub so we can't use them here. But as they
+	 * are limited to 32bpp RGB formats we don't actually need
+	 * to check anything.
+	 */
+	if (fb->modifier == I915_FORMAT_MOD_Y_TILED_CCS ||
+	    fb->modifier == I915_FORMAT_MOD_Yf_TILED_CCS)
+		return 0;
+
+	/*
 	 * Hardware doesn't handle subpixel coordinates.
 	 * Adjust to (macro)pixel boundary, but be careful not to
 	 * increase the source viewport size, because that could
@@ -297,26 +307,26 @@ int intel_plane_check_src_coordinates(struct intel_plane_state *plane_state)
 	drm_rect_init(src, src_x << 16, src_y << 16,
 		      src_w << 16, src_h << 16);
 
-	if (!fb->format->is_yuv)
-		return 0;
-
-	/* YUV specific checks */
-	if (!rotated) {
+	if (fb->format->format == DRM_FORMAT_RGB565 && rotated) {
+		hsub = 2;
+		vsub = 2;
+	} else {
 		hsub = fb->format->hsub;
 		vsub = fb->format->vsub;
-	} else {
-		hsub = vsub = max(fb->format->hsub, fb->format->vsub);
 	}
 
+	if (rotated)
+		hsub = vsub = max(hsub, vsub);
+
 	if (src_x % hsub || src_w % hsub) {
-		DRM_DEBUG_KMS("src x/w (%u, %u) must be a multiple of %u for %sYUV planes\n",
-			      src_x, src_w, hsub, rotated ? "rotated " : "");
+		DRM_DEBUG_KMS("src x/w (%u, %u) must be a multiple of %u (rotated: %s)\n",
+			      src_x, src_w, hsub, yesno(rotated));
 		return -EINVAL;
 	}
 
 	if (src_y % vsub || src_h % vsub) {
-		DRM_DEBUG_KMS("src y/h (%u, %u) must be a multiple of %u for %sYUV planes\n",
-			      src_y, src_h, vsub, rotated ? "rotated " : "");
+		DRM_DEBUG_KMS("src y/h (%u, %u) must be a multiple of %u (rotated: %s)\n",
+			      src_y, src_h, vsub, yesno(rotated));
 		return -EINVAL;
 	}
 
@@ -355,9 +365,8 @@ 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->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;
+	unsigned int pixel_rate = intel_plane_pixel_rate(crtc_state, plane_state);
 
 	skl_plane_ratio(crtc_state, plane_state, &num, &den);
 
@@ -365,17 +374,7 @@ 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->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);
-	dst_h = min(src_h, dst_h);
-
-	return DIV64_U64_ROUND_UP(mul_u32_u32(pixel_rate * num, src_w * src_h),
-				  mul_u32_u32(den, dst_w * dst_h));
+	return DIV_ROUND_UP(pixel_rate * num, den);
 }
 
 static unsigned int
@@ -2077,6 +2076,18 @@ vlv_sprite_check(struct intel_crtc_state *crtc_state,
 	return 0;
 }
 
+static bool intel_format_is_p01x(u32 format)
+{
+	switch (format) {
+	case DRM_FORMAT_P010:
+	case DRM_FORMAT_P012:
+	case DRM_FORMAT_P016:
+		return true;
+	default:
+		return false;
+	}
+}
+
 static int skl_plane_check_fb(const struct intel_crtc_state *crtc_state,
 			      const struct intel_plane_state *plane_state)
 {
@@ -2155,6 +2166,15 @@ static int skl_plane_check_fb(const struct intel_crtc_state *crtc_state,
 		return -EINVAL;
 	}
 
+	/* Wa_1606054188:tgl */
+	if (IS_TIGERLAKE(dev_priv) &&
+	    plane_state->ckey.flags & I915_SET_COLORKEY_SOURCE &&
+	    intel_format_is_p01x(fb->format->format)) {
+		drm_dbg_kms(&dev_priv->drm,
+			    "Source color keying not supported with P01x formats\n");
+		return -EINVAL;
+	}
+
 	return 0;
 }
 
@@ -3011,7 +3031,6 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 	struct intel_plane *plane;
 	enum drm_plane_type plane_type;
 	unsigned int supported_rotations;
-	unsigned int possible_crtcs;
 	const u64 *modifiers;
 	const u32 *formats;
 	int num_formats;
@@ -3066,10 +3085,8 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 	else
 		plane_type = DRM_PLANE_TYPE_OVERLAY;
 
-	possible_crtcs = BIT(pipe);
-
 	ret = drm_universal_plane_init(&dev_priv->drm, &plane->base,
-				       possible_crtcs, plane_funcs,
+				       0, plane_funcs,
 				       formats, num_formats, modifiers,
 				       plane_type,
 				       "plane %d%c", plane_id + 1,
@@ -3120,7 +3137,6 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 {
 	struct intel_plane *plane;
 	const struct drm_plane_funcs *plane_funcs;
-	unsigned long possible_crtcs;
 	unsigned int supported_rotations;
 	const u64 *modifiers;
 	const u32 *formats;
@@ -3205,10 +3221,8 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 	plane->id = PLANE_SPRITE0 + sprite;
 	plane->frontbuffer_bit = INTEL_FRONTBUFFER(pipe, plane->id);
 
-	possible_crtcs = BIT(pipe);
-
 	ret = drm_universal_plane_init(&dev_priv->drm, &plane->base,
-				       possible_crtcs, plane_funcs,
+				       0, plane_funcs,
 				       formats, num_formats, modifiers,
 				       DRM_PLANE_TYPE_OVERLAY,
 				       "sprite %c", sprite_name(pipe, sprite));
diff --git a/drivers/gpu/drm/i915/display/intel_tv.c b/drivers/gpu/drm/i915/display/intel_tv.c
index 4f81ee26b7ab..d2e3a3a323e9 100644
--- a/drivers/gpu/drm/i915/display/intel_tv.c
+++ b/drivers/gpu/drm/i915/display/intel_tv.c
@@ -33,7 +33,6 @@
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "intel_connector.h"
diff --git a/drivers/gpu/drm/i915/display/intel_vdsc.c b/drivers/gpu/drm/i915/display/intel_vdsc.c
index 7cba57ae72fe..95ad87d4ccb3 100644
--- a/drivers/gpu/drm/i915/display/intel_vdsc.c
+++ b/drivers/gpu/drm/i915/display/intel_vdsc.c
@@ -6,8 +6,6 @@
  *         Manasi Navare <manasi.d.navare@intel.com>
  */
 
-#include <drm/i915_drm.h>
-
 #include "i915_drv.h"
 #include "intel_display_types.h"
 #include "intel_dsi.h"
diff --git a/drivers/gpu/drm/i915/display/vlv_dsi.c b/drivers/gpu/drm/i915/display/vlv_dsi.c
index d07cfad8ce6f..f4c362dc6e15 100644
--- a/drivers/gpu/drm/i915/display/vlv_dsi.c
+++ b/drivers/gpu/drm/i915/display/vlv_dsi.c
@@ -1591,59 +1591,6 @@ static const struct drm_connector_funcs intel_dsi_connector_funcs = {
 	.atomic_duplicate_state = intel_digital_connector_duplicate_state,
 };
 
-static enum drm_panel_orientation
-vlv_dsi_get_hw_panel_orientation(struct intel_connector *connector)
-{
-	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
-	struct intel_encoder *encoder = intel_attached_encoder(connector);
-	enum intel_display_power_domain power_domain;
-	enum drm_panel_orientation orientation;
-	struct intel_plane *plane;
-	struct intel_crtc *crtc;
-	intel_wakeref_t wakeref;
-	enum pipe pipe;
-	u32 val;
-
-	if (!encoder->get_hw_state(encoder, &pipe))
-		return DRM_MODE_PANEL_ORIENTATION_UNKNOWN;
-
-	crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
-	plane = to_intel_plane(crtc->base.primary);
-
-	power_domain = POWER_DOMAIN_PIPE(pipe);
-	wakeref = intel_display_power_get_if_enabled(dev_priv, power_domain);
-	if (!wakeref)
-		return DRM_MODE_PANEL_ORIENTATION_UNKNOWN;
-
-	val = intel_de_read(dev_priv, DSPCNTR(plane->i9xx_plane));
-
-	if (!(val & DISPLAY_PLANE_ENABLE))
-		orientation = DRM_MODE_PANEL_ORIENTATION_UNKNOWN;
-	else if (val & DISPPLANE_ROTATE_180)
-		orientation = DRM_MODE_PANEL_ORIENTATION_BOTTOM_UP;
-	else
-		orientation = DRM_MODE_PANEL_ORIENTATION_NORMAL;
-
-	intel_display_power_put(dev_priv, power_domain, wakeref);
-
-	return orientation;
-}
-
-static enum drm_panel_orientation
-vlv_dsi_get_panel_orientation(struct intel_connector *connector)
-{
-	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
-	enum drm_panel_orientation orientation;
-
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
-		orientation = vlv_dsi_get_hw_panel_orientation(connector);
-		if (orientation != DRM_MODE_PANEL_ORIENTATION_UNKNOWN)
-			return orientation;
-	}
-
-	return intel_dsi_get_panel_orientation(connector);
-}
-
 static void vlv_dsi_add_properties(struct intel_connector *connector)
 {
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
@@ -1662,7 +1609,7 @@ static void vlv_dsi_add_properties(struct intel_connector *connector)
 
 		drm_connector_set_panel_orientation_with_quirk(
 				&connector->base,
-				vlv_dsi_get_panel_orientation(connector),
+				intel_dsi_get_panel_orientation(connector),
 				connector->panel.fixed_mode->hdisplay,
 				connector->panel.fixed_mode->vdisplay);
 	}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c
index 81366aa4812b..0598e5382a1d 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c
@@ -217,7 +217,7 @@ static void clear_pages_worker(struct work_struct *work)
 					   0);
 out_request:
 	if (unlikely(err)) {
-		i915_request_skip(rq, err);
+		i915_request_set_error_once(rq, err);
 		err = 0;
 	}
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c
index adcebf22a3d3..026999b34abd 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c
@@ -67,10 +67,9 @@
 #include <linux/log2.h>
 #include <linux/nospec.h>
 
-#include <drm/i915_drm.h>
-
 #include "gt/gen6_ppgtt.h"
 #include "gt/intel_context.h"
+#include "gt/intel_context_param.h"
 #include "gt/intel_engine_heartbeat.h"
 #include "gt/intel_engine_user.h"
 #include "gt/intel_ring.h"
@@ -243,7 +242,6 @@ static void __free_engines(struct i915_gem_engines *e, unsigned int count)
 		if (!e->engines[count])
 			continue;
 
-		RCU_INIT_POINTER(e->engines[count]->gem_context, NULL);
 		intel_context_put(e->engines[count]);
 	}
 	kfree(e);
@@ -256,7 +254,51 @@ static void free_engines(struct i915_gem_engines *e)
 
 static void free_engines_rcu(struct rcu_head *rcu)
 {
-	free_engines(container_of(rcu, struct i915_gem_engines, rcu));
+	struct i915_gem_engines *engines =
+		container_of(rcu, struct i915_gem_engines, rcu);
+
+	i915_sw_fence_fini(&engines->fence);
+	free_engines(engines);
+}
+
+static int __i915_sw_fence_call
+engines_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
+{
+	struct i915_gem_engines *engines =
+		container_of(fence, typeof(*engines), fence);
+
+	switch (state) {
+	case FENCE_COMPLETE:
+		if (!list_empty(&engines->link)) {
+			struct i915_gem_context *ctx = engines->ctx;
+			unsigned long flags;
+
+			spin_lock_irqsave(&ctx->stale.lock, flags);
+			list_del(&engines->link);
+			spin_unlock_irqrestore(&ctx->stale.lock, flags);
+		}
+		i915_gem_context_put(engines->ctx);
+		break;
+
+	case FENCE_FREE:
+		init_rcu_head(&engines->rcu);
+		call_rcu(&engines->rcu, free_engines_rcu);
+		break;
+	}
+
+	return NOTIFY_DONE;
+}
+
+static struct i915_gem_engines *alloc_engines(unsigned int count)
+{
+	struct i915_gem_engines *e;
+
+	e = kzalloc(struct_size(e, engines, count), GFP_KERNEL);
+	if (!e)
+		return NULL;
+
+	i915_sw_fence_init(&e->fence, engines_notify);
+	return e;
 }
 
 static struct i915_gem_engines *default_engines(struct i915_gem_context *ctx)
@@ -266,12 +308,10 @@ static struct i915_gem_engines *default_engines(struct i915_gem_context *ctx)
 	struct i915_gem_engines *e;
 	enum intel_engine_id id;
 
-	e = kzalloc(struct_size(e, engines, I915_NUM_ENGINES), GFP_KERNEL);
+	e = alloc_engines(I915_NUM_ENGINES);
 	if (!e)
 		return ERR_PTR(-ENOMEM);
 
-	e->ctx = ctx;
-
 	for_each_engine(engine, gt, id) {
 		struct intel_context *ce;
 
@@ -305,7 +345,6 @@ static void i915_gem_context_free(struct i915_gem_context *ctx)
 	list_del(&ctx->link);
 	spin_unlock(&ctx->i915->gem.contexts.lock);
 
-	free_engines(rcu_access_pointer(ctx->engines));
 	mutex_destroy(&ctx->engines_mutex);
 
 	if (ctx->timeline)
@@ -492,30 +531,75 @@ static void kill_engines(struct i915_gem_engines *engines)
 static void kill_stale_engines(struct i915_gem_context *ctx)
 {
 	struct i915_gem_engines *pos, *next;
-	unsigned long flags;
 
-	spin_lock_irqsave(&ctx->stale.lock, flags);
+	spin_lock_irq(&ctx->stale.lock);
+	GEM_BUG_ON(!i915_gem_context_is_closed(ctx));
 	list_for_each_entry_safe(pos, next, &ctx->stale.engines, link) {
-		if (!i915_sw_fence_await(&pos->fence))
+		if (!i915_sw_fence_await(&pos->fence)) {
+			list_del_init(&pos->link);
 			continue;
+		}
 
-		spin_unlock_irqrestore(&ctx->stale.lock, flags);
+		spin_unlock_irq(&ctx->stale.lock);
 
 		kill_engines(pos);
 
-		spin_lock_irqsave(&ctx->stale.lock, flags);
+		spin_lock_irq(&ctx->stale.lock);
+		GEM_BUG_ON(i915_sw_fence_signaled(&pos->fence));
 		list_safe_reset_next(pos, next, link);
 		list_del_init(&pos->link); /* decouple from FENCE_COMPLETE */
 
 		i915_sw_fence_complete(&pos->fence);
 	}
-	spin_unlock_irqrestore(&ctx->stale.lock, flags);
+	spin_unlock_irq(&ctx->stale.lock);
 }
 
 static void kill_context(struct i915_gem_context *ctx)
 {
 	kill_stale_engines(ctx);
-	kill_engines(__context_engines_static(ctx));
+}
+
+static void engines_idle_release(struct i915_gem_context *ctx,
+				 struct i915_gem_engines *engines)
+{
+	struct i915_gem_engines_iter it;
+	struct intel_context *ce;
+
+	INIT_LIST_HEAD(&engines->link);
+
+	engines->ctx = i915_gem_context_get(ctx);
+
+	for_each_gem_engine(ce, engines, it) {
+		struct dma_fence *fence;
+		int err = 0;
+
+		/* serialises with execbuf */
+		RCU_INIT_POINTER(ce->gem_context, NULL);
+		if (!intel_context_pin_if_active(ce))
+			continue;
+
+		fence = i915_active_fence_get(&ce->timeline->last_request);
+		if (fence) {
+			err = i915_sw_fence_await_dma_fence(&engines->fence,
+							    fence, 0,
+							    GFP_KERNEL);
+			dma_fence_put(fence);
+		}
+		intel_context_unpin(ce);
+		if (err < 0)
+			goto kill;
+	}
+
+	spin_lock_irq(&ctx->stale.lock);
+	if (!i915_gem_context_is_closed(ctx))
+		list_add_tail(&engines->link, &ctx->stale.engines);
+	spin_unlock_irq(&ctx->stale.lock);
+
+kill:
+	if (list_empty(&engines->link)) /* raced, already closed */
+		kill_engines(engines);
+
+	i915_sw_fence_commit(&engines->fence);
 }
 
 static void set_closed_name(struct i915_gem_context *ctx)
@@ -539,11 +623,16 @@ static void context_close(struct i915_gem_context *ctx)
 {
 	struct i915_address_space *vm;
 
+	/* Flush any concurrent set_engines() */
+	mutex_lock(&ctx->engines_mutex);
+	engines_idle_release(ctx, rcu_replace_pointer(ctx->engines, NULL, 1));
 	i915_gem_context_set_closed(ctx);
-	set_closed_name(ctx);
+	mutex_unlock(&ctx->engines_mutex);
 
 	mutex_lock(&ctx->mutex);
 
+	set_closed_name(ctx);
+
 	vm = i915_gem_context_vm(ctx);
 	if (vm)
 		i915_vm_close(vm);
@@ -668,23 +757,30 @@ err_free:
 	return ERR_PTR(err);
 }
 
-static void
+static int
 context_apply_all(struct i915_gem_context *ctx,
-		  void (*fn)(struct intel_context *ce, void *data),
+		  int (*fn)(struct intel_context *ce, void *data),
 		  void *data)
 {
 	struct i915_gem_engines_iter it;
 	struct intel_context *ce;
+	int err = 0;
 
-	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it)
-		fn(ce, data);
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+		err = fn(ce, data);
+		if (err)
+			break;
+	}
 	i915_gem_context_unlock_engines(ctx);
+
+	return err;
 }
 
-static void __apply_ppgtt(struct intel_context *ce, void *vm)
+static int __apply_ppgtt(struct intel_context *ce, void *vm)
 {
 	i915_vm_put(ce->vm);
 	ce->vm = i915_vm_get(vm);
+	return 0;
 }
 
 static struct i915_address_space *
@@ -722,9 +818,10 @@ static void __set_timeline(struct intel_timeline **dst,
 		intel_timeline_put(old);
 }
 
-static void __apply_timeline(struct intel_context *ce, void *timeline)
+static int __apply_timeline(struct intel_context *ce, void *timeline)
 {
 	__set_timeline(&ce->timeline, timeline);
+	return 0;
 }
 
 static void __assign_timeline(struct i915_gem_context *ctx,
@@ -806,6 +903,7 @@ void i915_gem_init__contexts(struct drm_i915_private *i915)
 void i915_gem_driver_release__contexts(struct drm_i915_private *i915)
 {
 	flush_work(&i915->gem.contexts.free_work);
+	rcu_barrier(); /* and flush the left over RCU frees */
 }
 
 static int gem_context_register(struct i915_gem_context *ctx,
@@ -971,6 +1069,30 @@ static void cb_retire(struct i915_active *base)
 	kfree(cb);
 }
 
+static inline struct i915_gem_engines *
+__context_engines_await(const struct i915_gem_context *ctx)
+{
+	struct i915_gem_engines *engines;
+
+	rcu_read_lock();
+	do {
+		engines = rcu_dereference(ctx->engines);
+		if (unlikely(!engines))
+			break;
+
+		if (unlikely(!i915_sw_fence_await(&engines->fence)))
+			continue;
+
+		if (likely(engines == rcu_access_pointer(ctx->engines)))
+			break;
+
+		i915_sw_fence_complete(&engines->fence);
+	} while (1);
+	rcu_read_unlock();
+
+	return engines;
+}
+
 I915_SELFTEST_DECLARE(static intel_engine_mask_t context_barrier_inject_fault);
 static int context_barrier_task(struct i915_gem_context *ctx,
 				intel_engine_mask_t engines,
@@ -981,6 +1103,7 @@ static int context_barrier_task(struct i915_gem_context *ctx,
 {
 	struct context_barrier_task *cb;
 	struct i915_gem_engines_iter it;
+	struct i915_gem_engines *e;
 	struct intel_context *ce;
 	int err = 0;
 
@@ -997,7 +1120,13 @@ static int context_barrier_task(struct i915_gem_context *ctx,
 		return err;
 	}
 
-	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+	e = __context_engines_await(ctx);
+	if (!e) {
+		i915_active_release(&cb->base);
+		return -ENOENT;
+	}
+
+	for_each_gem_engine(ce, e, it) {
 		struct i915_request *rq;
 
 		if (I915_SELFTEST_ONLY(context_barrier_inject_fault &
@@ -1028,7 +1157,7 @@ static int context_barrier_task(struct i915_gem_context *ctx,
 		if (err)
 			break;
 	}
-	i915_gem_context_unlock_engines(ctx);
+	i915_sw_fence_complete(&e->fence);
 
 	cb->task = err ? NULL : task; /* caller needs to unwind instead */
 	cb->data = data;
@@ -1215,6 +1344,63 @@ out:
 	return err;
 }
 
+static int __apply_ringsize(struct intel_context *ce, void *sz)
+{
+	return intel_context_set_ring_size(ce, (unsigned long)sz);
+}
+
+static int set_ringsize(struct i915_gem_context *ctx,
+			struct drm_i915_gem_context_param *args)
+{
+	if (!HAS_LOGICAL_RING_CONTEXTS(ctx->i915))
+		return -ENODEV;
+
+	if (args->size)
+		return -EINVAL;
+
+	if (!IS_ALIGNED(args->value, I915_GTT_PAGE_SIZE))
+		return -EINVAL;
+
+	if (args->value < I915_GTT_PAGE_SIZE)
+		return -EINVAL;
+
+	if (args->value > 128 * I915_GTT_PAGE_SIZE)
+		return -EINVAL;
+
+	return context_apply_all(ctx,
+				 __apply_ringsize,
+				 __intel_context_ring_size(args->value));
+}
+
+static int __get_ringsize(struct intel_context *ce, void *arg)
+{
+	long sz;
+
+	sz = intel_context_get_ring_size(ce);
+	GEM_BUG_ON(sz > INT_MAX);
+
+	return sz; /* stop on first engine */
+}
+
+static int get_ringsize(struct i915_gem_context *ctx,
+			struct drm_i915_gem_context_param *args)
+{
+	int sz;
+
+	if (!HAS_LOGICAL_RING_CONTEXTS(ctx->i915))
+		return -ENODEV;
+
+	if (args->size)
+		return -EINVAL;
+
+	sz = context_apply_all(ctx, __get_ringsize, NULL);
+	if (sz < 0)
+		return sz;
+
+	args->value = sz;
+	return 0;
+}
+
 static int
 user_to_context_sseu(struct drm_i915_private *i915,
 		     const struct drm_i915_gem_context_param_sseu *user,
@@ -1562,77 +1748,6 @@ static const i915_user_extension_fn set_engines__extensions[] = {
 	[I915_CONTEXT_ENGINES_EXT_BOND] = set_engines__bond,
 };
 
-static int engines_notify(struct i915_sw_fence *fence,
-			  enum i915_sw_fence_notify state)
-{
-	struct i915_gem_engines *engines =
-		container_of(fence, typeof(*engines), fence);
-
-	switch (state) {
-	case FENCE_COMPLETE:
-		if (!list_empty(&engines->link)) {
-			struct i915_gem_context *ctx = engines->ctx;
-			unsigned long flags;
-
-			spin_lock_irqsave(&ctx->stale.lock, flags);
-			list_del(&engines->link);
-			spin_unlock_irqrestore(&ctx->stale.lock, flags);
-		}
-		break;
-
-	case FENCE_FREE:
-		init_rcu_head(&engines->rcu);
-		call_rcu(&engines->rcu, free_engines_rcu);
-		break;
-	}
-
-	return NOTIFY_DONE;
-}
-
-static void engines_idle_release(struct i915_gem_engines *engines)
-{
-	struct i915_gem_engines_iter it;
-	struct intel_context *ce;
-	unsigned long flags;
-
-	GEM_BUG_ON(!engines);
-	i915_sw_fence_init(&engines->fence, engines_notify);
-
-	INIT_LIST_HEAD(&engines->link);
-	spin_lock_irqsave(&engines->ctx->stale.lock, flags);
-	if (!i915_gem_context_is_closed(engines->ctx))
-		list_add(&engines->link, &engines->ctx->stale.engines);
-	spin_unlock_irqrestore(&engines->ctx->stale.lock, flags);
-	if (list_empty(&engines->link)) /* raced, already closed */
-		goto kill;
-
-	for_each_gem_engine(ce, engines, it) {
-		struct dma_fence *fence;
-		int err;
-
-		if (!ce->timeline)
-			continue;
-
-		fence = i915_active_fence_get(&ce->timeline->last_request);
-		if (!fence)
-			continue;
-
-		err = i915_sw_fence_await_dma_fence(&engines->fence,
-						    fence, 0,
-						    GFP_KERNEL);
-
-		dma_fence_put(fence);
-		if (err < 0)
-			goto kill;
-	}
-	goto out;
-
-kill:
-	kill_engines(engines);
-out:
-	i915_sw_fence_commit(&engines->fence);
-}
-
 static int
 set_engines(struct i915_gem_context *ctx,
 	    const struct drm_i915_gem_context_param *args)
@@ -1669,14 +1784,10 @@ set_engines(struct i915_gem_context *ctx,
 	 * first 64 engines defined here.
 	 */
 	num_engines = (args->size - sizeof(*user)) / sizeof(*user->engines);
-
-	set.engines = kmalloc(struct_size(set.engines, engines, num_engines),
-			      GFP_KERNEL);
+	set.engines = alloc_engines(num_engines);
 	if (!set.engines)
 		return -ENOMEM;
 
-	set.engines->ctx = ctx;
-
 	for (n = 0; n < num_engines; n++) {
 		struct i915_engine_class_instance ci;
 		struct intel_engine_cs *engine;
@@ -1729,6 +1840,11 @@ set_engines(struct i915_gem_context *ctx,
 
 replace:
 	mutex_lock(&ctx->engines_mutex);
+	if (i915_gem_context_is_closed(ctx)) {
+		mutex_unlock(&ctx->engines_mutex);
+		free_engines(set.engines);
+		return -ENOENT;
+	}
 	if (args->size)
 		i915_gem_context_set_user_engines(ctx);
 	else
@@ -1737,7 +1853,7 @@ replace:
 	mutex_unlock(&ctx->engines_mutex);
 
 	/* Keep track of old engine sets for kill_context() */
-	engines_idle_release(set.engines);
+	engines_idle_release(ctx, set.engines);
 
 	return 0;
 }
@@ -1748,7 +1864,7 @@ __copy_engines(struct i915_gem_engines *e)
 	struct i915_gem_engines *copy;
 	unsigned int n;
 
-	copy = kmalloc(struct_size(e, engines, e->num_engines), GFP_KERNEL);
+	copy = alloc_engines(e->num_engines);
 	if (!copy)
 		return ERR_PTR(-ENOMEM);
 
@@ -1852,17 +1968,19 @@ set_persistence(struct i915_gem_context *ctx,
 	return __context_set_persistence(ctx, args->value);
 }
 
-static void __apply_priority(struct intel_context *ce, void *arg)
+static int __apply_priority(struct intel_context *ce, void *arg)
 {
 	struct i915_gem_context *ctx = arg;
 
 	if (!intel_engine_has_semaphores(ce->engine))
-		return;
+		return 0;
 
 	if (ctx->sched.priority >= I915_PRIORITY_NORMAL)
 		intel_context_set_use_semaphores(ce);
 	else
 		intel_context_clear_use_semaphores(ce);
+
+	return 0;
 }
 
 static int set_priority(struct i915_gem_context *ctx,
@@ -1955,6 +2073,10 @@ static int ctx_setparam(struct drm_i915_file_private *fpriv,
 		ret = set_persistence(ctx, args);
 		break;
 
+	case I915_CONTEXT_PARAM_RINGSIZE:
+		ret = set_ringsize(ctx, args);
+		break;
+
 	case I915_CONTEXT_PARAM_BAN_PERIOD:
 	default:
 		ret = -EINVAL;
@@ -1983,6 +2105,18 @@ static int create_setparam(struct i915_user_extension __user *ext, void *data)
 	return ctx_setparam(arg->fpriv, arg->ctx, &local.param);
 }
 
+static int copy_ring_size(struct intel_context *dst,
+			  struct intel_context *src)
+{
+	long sz;
+
+	sz = intel_context_get_ring_size(src);
+	if (sz < 0)
+		return sz;
+
+	return intel_context_set_ring_size(dst, sz);
+}
+
 static int clone_engines(struct i915_gem_context *dst,
 			 struct i915_gem_context *src)
 {
@@ -1991,12 +2125,10 @@ static int clone_engines(struct i915_gem_context *dst,
 	bool user_engines;
 	unsigned long n;
 
-	clone = kmalloc(struct_size(e, engines, e->num_engines), GFP_KERNEL);
+	clone = alloc_engines(e->num_engines);
 	if (!clone)
 		goto err_unlock;
 
-	clone->ctx = dst;
-
 	for (n = 0; n < e->num_engines; n++) {
 		struct intel_engine_cs *engine;
 
@@ -2026,6 +2158,12 @@ static int clone_engines(struct i915_gem_context *dst,
 		}
 
 		intel_context_set_gem(clone->engines[n], dst);
+
+		/* Copy across the preferred ringsize */
+		if (copy_ring_size(clone->engines[n], e->engines[n])) {
+			__free_engines(clone, n + 1);
+			goto err_unlock;
+		}
 	}
 	clone->num_engines = n;
 
@@ -2033,8 +2171,7 @@ static int clone_engines(struct i915_gem_context *dst,
 	i915_gem_context_unlock_engines(src);
 
 	/* Serialised by constructor */
-	free_engines(__context_engines_static(dst));
-	RCU_INIT_POINTER(dst->engines, clone);
+	engines_idle_release(dst, rcu_replace_pointer(dst->engines, clone, 1));
 	if (user_engines)
 		i915_gem_context_set_user_engines(dst);
 	else
@@ -2388,6 +2525,10 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data,
 		args->value = i915_gem_context_is_persistent(ctx);
 		break;
 
+	case I915_CONTEXT_PARAM_RINGSIZE:
+		ret = get_ringsize(ctx, args);
+		break;
+
 	case I915_CONTEXT_PARAM_BAN_PERIOD:
 	default:
 		ret = -EINVAL;
@@ -2461,6 +2602,9 @@ i915_gem_engines_iter_next(struct i915_gem_engines_iter *it)
 	const struct i915_gem_engines *e = it->engines;
 	struct intel_context *ctx;
 
+	if (unlikely(!e))
+		return NULL;
+
 	do {
 		if (it->idx >= e->num_engines)
 			return NULL;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.h b/drivers/gpu/drm/i915/gem/i915_gem_context.h
index 3ae61a355d87..57b7ae2893e1 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.h
@@ -207,7 +207,6 @@ static inline void
 i915_gem_engines_iter_init(struct i915_gem_engines_iter *it,
 			   struct i915_gem_engines *engines)
 {
-	GEM_BUG_ON(!engines);
 	it->engines = engines;
 	it->idx = 0;
 }
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
index 4f9c1f5a4ded..d3f4f28e9468 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
@@ -10,7 +10,6 @@
 #include <linux/uaccess.h>
 
 #include <drm/drm_syncobj.h>
-#include <drm/i915_drm.h>
 
 #include "display/intel_frontbuffer.h"
 
@@ -28,6 +27,19 @@
 #include "i915_sw_fence_work.h"
 #include "i915_trace.h"
 
+struct eb_vma {
+	struct i915_vma *vma;
+	unsigned int flags;
+
+	/** This vma's place in the execbuf reservation list */
+	struct drm_i915_gem_exec_object2 *exec;
+	struct list_head bind_link;
+	struct list_head reloc_link;
+
+	struct hlist_node node;
+	u32 handle;
+};
+
 enum {
 	FORCE_CPU_RELOC = 1,
 	FORCE_GTT_RELOC,
@@ -35,17 +47,15 @@ enum {
 #define DBG_FORCE_RELOC 0 /* choose one of the above! */
 };
 
-#define __EXEC_OBJECT_HAS_REF		BIT(31)
-#define __EXEC_OBJECT_HAS_PIN		BIT(30)
-#define __EXEC_OBJECT_HAS_FENCE		BIT(29)
-#define __EXEC_OBJECT_NEEDS_MAP		BIT(28)
-#define __EXEC_OBJECT_NEEDS_BIAS	BIT(27)
-#define __EXEC_OBJECT_INTERNAL_FLAGS	(~0u << 27) /* all of the above */
+#define __EXEC_OBJECT_HAS_PIN		BIT(31)
+#define __EXEC_OBJECT_HAS_FENCE		BIT(30)
+#define __EXEC_OBJECT_NEEDS_MAP		BIT(29)
+#define __EXEC_OBJECT_NEEDS_BIAS	BIT(28)
+#define __EXEC_OBJECT_INTERNAL_FLAGS	(~0u << 28) /* all of the above */
 #define __EXEC_OBJECT_RESERVED (__EXEC_OBJECT_HAS_PIN | __EXEC_OBJECT_HAS_FENCE)
 
 #define __EXEC_HAS_RELOC	BIT(31)
-#define __EXEC_VALIDATED	BIT(30)
-#define __EXEC_INTERNAL_FLAGS	(~0u << 30)
+#define __EXEC_INTERNAL_FLAGS	(~0u << 31)
 #define UPDATE			PIN_OFFSET_FIXED
 
 #define BATCH_OFFSET_BIAS (256*1024)
@@ -220,15 +230,14 @@ struct i915_execbuffer {
 	struct drm_file *file; /** per-file lookup tables and limits */
 	struct drm_i915_gem_execbuffer2 *args; /** ioctl parameters */
 	struct drm_i915_gem_exec_object2 *exec; /** ioctl execobj[] */
-	struct i915_vma **vma;
-	unsigned int *flags;
+	struct eb_vma *vma;
 
 	struct intel_engine_cs *engine; /** engine to queue the request to */
 	struct intel_context *context; /* logical state for the request */
 	struct i915_gem_context *gem_context; /** caller's context */
 
 	struct i915_request *request; /** our request to build */
-	struct i915_vma *batch; /** identity of the batch obj/vma */
+	struct eb_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 */
@@ -276,8 +285,6 @@ struct i915_execbuffer {
 	struct hlist_head *buckets; /** ht for relocation handles */
 };
 
-#define exec_entry(EB, VMA) (&(EB)->exec[(VMA)->exec_flags - (EB)->flags])
-
 static inline bool eb_use_cmdparser(const struct i915_execbuffer *eb)
 {
 	return intel_engine_requires_cmd_parser(eb->engine) ||
@@ -364,9 +371,9 @@ eb_vma_misplaced(const struct drm_i915_gem_exec_object2 *entry,
 static inline bool
 eb_pin_vma(struct i915_execbuffer *eb,
 	   const struct drm_i915_gem_exec_object2 *entry,
-	   struct i915_vma *vma)
+	   struct eb_vma *ev)
 {
-	unsigned int exec_flags = *vma->exec_flags;
+	struct i915_vma *vma = ev->vma;
 	u64 pin_flags;
 
 	if (vma->node.size)
@@ -375,24 +382,24 @@ eb_pin_vma(struct i915_execbuffer *eb,
 		pin_flags = entry->offset & PIN_OFFSET_MASK;
 
 	pin_flags |= PIN_USER | PIN_NOEVICT | PIN_OFFSET_FIXED;
-	if (unlikely(exec_flags & EXEC_OBJECT_NEEDS_GTT))
+	if (unlikely(ev->flags & EXEC_OBJECT_NEEDS_GTT))
 		pin_flags |= PIN_GLOBAL;
 
 	if (unlikely(i915_vma_pin(vma, 0, 0, pin_flags)))
 		return false;
 
-	if (unlikely(exec_flags & EXEC_OBJECT_NEEDS_FENCE)) {
+	if (unlikely(ev->flags & EXEC_OBJECT_NEEDS_FENCE)) {
 		if (unlikely(i915_vma_pin_fence(vma))) {
 			i915_vma_unpin(vma);
 			return false;
 		}
 
 		if (vma->fence)
-			exec_flags |= __EXEC_OBJECT_HAS_FENCE;
+			ev->flags |= __EXEC_OBJECT_HAS_FENCE;
 	}
 
-	*vma->exec_flags = exec_flags | __EXEC_OBJECT_HAS_PIN;
-	return !eb_vma_misplaced(entry, vma, exec_flags);
+	ev->flags |= __EXEC_OBJECT_HAS_PIN;
+	return !eb_vma_misplaced(entry, vma, ev->flags);
 }
 
 static inline void __eb_unreserve_vma(struct i915_vma *vma, unsigned int flags)
@@ -406,13 +413,13 @@ static inline void __eb_unreserve_vma(struct i915_vma *vma, unsigned int flags)
 }
 
 static inline void
-eb_unreserve_vma(struct i915_vma *vma, unsigned int *flags)
+eb_unreserve_vma(struct eb_vma *ev)
 {
-	if (!(*flags & __EXEC_OBJECT_HAS_PIN))
+	if (!(ev->flags & __EXEC_OBJECT_HAS_PIN))
 		return;
 
-	__eb_unreserve_vma(vma, *flags);
-	*flags &= ~__EXEC_OBJECT_RESERVED;
+	__eb_unreserve_vma(ev->vma, ev->flags);
+	ev->flags &= ~__EXEC_OBJECT_RESERVED;
 }
 
 static int
@@ -420,11 +427,11 @@ eb_validate_vma(struct i915_execbuffer *eb,
 		struct drm_i915_gem_exec_object2 *entry,
 		struct i915_vma *vma)
 {
-	struct drm_i915_private *i915 = eb->i915;
 	if (unlikely(entry->flags & eb->invalid_flags))
 		return -EINVAL;
 
-	if (unlikely(entry->alignment && !is_power_of_2(entry->alignment)))
+	if (unlikely(entry->alignment &&
+		     !is_power_of_2_u64(entry->alignment)))
 		return -EINVAL;
 
 	/*
@@ -442,14 +449,6 @@ eb_validate_vma(struct i915_execbuffer *eb,
 	} else {
 		entry->pad_to_size = 0;
 	}
-
-	if (unlikely(vma->exec_flags)) {
-		drm_dbg(&i915->drm,
-			"Object [handle %d, index %d] appears more than once in object list\n",
-			entry->handle, (int)(entry - eb->exec));
-		return -EINVAL;
-	}
-
 	/*
 	 * From drm_mm perspective address space is continuous,
 	 * so from this point we're always using non-canonical
@@ -472,41 +471,29 @@ eb_validate_vma(struct i915_execbuffer *eb,
 	return 0;
 }
 
-static int
+static void
 eb_add_vma(struct i915_execbuffer *eb,
 	   unsigned int i, unsigned batch_idx,
 	   struct i915_vma *vma)
 {
 	struct drm_i915_gem_exec_object2 *entry = &eb->exec[i];
-	int err;
+	struct eb_vma *ev = &eb->vma[i];
 
 	GEM_BUG_ON(i915_vma_is_closed(vma));
 
-	if (!(eb->args->flags & __EXEC_VALIDATED)) {
-		err = eb_validate_vma(eb, entry, vma);
-		if (unlikely(err))
-			return err;
-	}
+	ev->vma = i915_vma_get(vma);
+	ev->exec = entry;
+	ev->flags = entry->flags;
 
 	if (eb->lut_size > 0) {
-		vma->exec_handle = entry->handle;
-		hlist_add_head(&vma->exec_node,
+		ev->handle = entry->handle;
+		hlist_add_head(&ev->node,
 			       &eb->buckets[hash_32(entry->handle,
 						    eb->lut_size)]);
 	}
 
 	if (entry->relocation_count)
-		list_add_tail(&vma->reloc_link, &eb->relocs);
-
-	/*
-	 * Stash a pointer from the vma to execobj, so we can query its flags,
-	 * size, alignment etc as provided by the user. Also we stash a pointer
-	 * to the vma inside the execobj so that we can use a direct lookup
-	 * to find the right target VMA when doing relocations.
-	 */
-	eb->vma[i] = vma;
-	eb->flags[i] = entry->flags;
-	vma->exec_flags = &eb->flags[i];
+		list_add_tail(&ev->reloc_link, &eb->relocs);
 
 	/*
 	 * SNA is doing fancy tricks with compressing batch buffers, which leads
@@ -519,30 +506,23 @@ eb_add_vma(struct i915_execbuffer *eb,
 	 */
 	if (i == batch_idx) {
 		if (entry->relocation_count &&
-		    !(eb->flags[i] & EXEC_OBJECT_PINNED))
-			eb->flags[i] |= __EXEC_OBJECT_NEEDS_BIAS;
+		    !(ev->flags & EXEC_OBJECT_PINNED))
+			ev->flags |= __EXEC_OBJECT_NEEDS_BIAS;
 		if (eb->reloc_cache.has_fence)
-			eb->flags[i] |= EXEC_OBJECT_NEEDS_FENCE;
+			ev->flags |= EXEC_OBJECT_NEEDS_FENCE;
 
-		eb->batch = vma;
+		eb->batch = ev;
 	}
 
-	err = 0;
-	if (eb_pin_vma(eb, entry, vma)) {
+	if (eb_pin_vma(eb, entry, ev)) {
 		if (entry->offset != vma->node.start) {
 			entry->offset = vma->node.start | UPDATE;
 			eb->args->flags |= __EXEC_HAS_RELOC;
 		}
 	} else {
-		eb_unreserve_vma(vma, vma->exec_flags);
-
-		list_add_tail(&vma->exec_link, &eb->unbound);
-		if (drm_mm_node_allocated(&vma->node))
-			err = i915_vma_unbind(vma);
-		if (unlikely(err))
-			vma->exec_flags = NULL;
+		eb_unreserve_vma(ev);
+		list_add_tail(&ev->bind_link, &eb->unbound);
 	}
-	return err;
 }
 
 static inline int use_cpu_reloc(const struct reloc_cache *cache,
@@ -563,14 +543,14 @@ static inline int use_cpu_reloc(const struct reloc_cache *cache,
 }
 
 static int eb_reserve_vma(const struct i915_execbuffer *eb,
-			  struct i915_vma *vma)
+			  struct eb_vma *ev,
+			  u64 pin_flags)
 {
-	struct drm_i915_gem_exec_object2 *entry = exec_entry(eb, vma);
-	unsigned int exec_flags = *vma->exec_flags;
-	u64 pin_flags;
+	struct drm_i915_gem_exec_object2 *entry = ev->exec;
+	unsigned int exec_flags = ev->flags;
+	struct i915_vma *vma = ev->vma;
 	int err;
 
-	pin_flags = PIN_USER | PIN_NONBLOCK;
 	if (exec_flags & EXEC_OBJECT_NEEDS_GTT)
 		pin_flags |= PIN_GLOBAL;
 
@@ -584,11 +564,16 @@ static int eb_reserve_vma(const struct i915_execbuffer *eb,
 	if (exec_flags & __EXEC_OBJECT_NEEDS_MAP)
 		pin_flags |= PIN_MAPPABLE;
 
-	if (exec_flags & EXEC_OBJECT_PINNED) {
+	if (exec_flags & EXEC_OBJECT_PINNED)
 		pin_flags |= entry->offset | PIN_OFFSET_FIXED;
-		pin_flags &= ~PIN_NONBLOCK; /* force overlapping checks */
-	} else if (exec_flags & __EXEC_OBJECT_NEEDS_BIAS) {
+	else if (exec_flags & __EXEC_OBJECT_NEEDS_BIAS)
 		pin_flags |= BATCH_OFFSET_BIAS | PIN_OFFSET_BIAS;
+
+	if (drm_mm_node_allocated(&vma->node) &&
+	    eb_vma_misplaced(entry, vma, ev->flags)) {
+		err = i915_vma_unbind(vma);
+		if (err)
+			return err;
 	}
 
 	err = i915_vma_pin(vma,
@@ -613,8 +598,8 @@ static int eb_reserve_vma(const struct i915_execbuffer *eb,
 			exec_flags |= __EXEC_OBJECT_HAS_FENCE;
 	}
 
-	*vma->exec_flags = exec_flags | __EXEC_OBJECT_HAS_PIN;
-	GEM_BUG_ON(eb_vma_misplaced(entry, vma, exec_flags));
+	ev->flags = exec_flags | __EXEC_OBJECT_HAS_PIN;
+	GEM_BUG_ON(eb_vma_misplaced(entry, vma, ev->flags));
 
 	return 0;
 }
@@ -622,10 +607,11 @@ static int eb_reserve_vma(const struct i915_execbuffer *eb,
 static int eb_reserve(struct i915_execbuffer *eb)
 {
 	const unsigned int count = eb->buffer_count;
+	unsigned int pin_flags = PIN_USER | PIN_NONBLOCK;
 	struct list_head last;
-	struct i915_vma *vma;
+	struct eb_vma *ev;
 	unsigned int i, pass;
-	int err;
+	int err = 0;
 
 	/*
 	 * Attempt to pin all of the buffers into the GTT.
@@ -641,44 +627,54 @@ static int eb_reserve(struct i915_execbuffer *eb)
 	 * room for the earlier objects *unless* we need to defragment.
 	 */
 
+	if (mutex_lock_interruptible(&eb->i915->drm.struct_mutex))
+		return -EINTR;
+
 	pass = 0;
-	err = 0;
 	do {
-		list_for_each_entry(vma, &eb->unbound, exec_link) {
-			err = eb_reserve_vma(eb, vma);
+		list_for_each_entry(ev, &eb->unbound, bind_link) {
+			err = eb_reserve_vma(eb, ev, pin_flags);
 			if (err)
 				break;
 		}
-		if (err != -ENOSPC)
-			return err;
+		if (!(err == -ENOSPC || err == -EAGAIN))
+			break;
 
 		/* Resort *all* the objects into priority order */
 		INIT_LIST_HEAD(&eb->unbound);
 		INIT_LIST_HEAD(&last);
 		for (i = 0; i < count; i++) {
-			unsigned int flags = eb->flags[i];
-			struct i915_vma *vma = eb->vma[i];
+			unsigned int flags;
 
+			ev = &eb->vma[i];
+			flags = ev->flags;
 			if (flags & EXEC_OBJECT_PINNED &&
 			    flags & __EXEC_OBJECT_HAS_PIN)
 				continue;
 
-			eb_unreserve_vma(vma, &eb->flags[i]);
+			eb_unreserve_vma(ev);
 
 			if (flags & EXEC_OBJECT_PINNED)
 				/* Pinned must have their slot */
-				list_add(&vma->exec_link, &eb->unbound);
+				list_add(&ev->bind_link, &eb->unbound);
 			else if (flags & __EXEC_OBJECT_NEEDS_MAP)
 				/* Map require the lowest 256MiB (aperture) */
-				list_add_tail(&vma->exec_link, &eb->unbound);
+				list_add_tail(&ev->bind_link, &eb->unbound);
 			else if (!(flags & EXEC_OBJECT_SUPPORTS_48B_ADDRESS))
 				/* Prioritise 4GiB region for restricted bo */
-				list_add(&vma->exec_link, &last);
+				list_add(&ev->bind_link, &last);
 			else
-				list_add_tail(&vma->exec_link, &last);
+				list_add_tail(&ev->bind_link, &last);
 		}
 		list_splice_tail(&last, &eb->unbound);
 
+		if (err == -EAGAIN) {
+			mutex_unlock(&eb->i915->drm.struct_mutex);
+			flush_workqueue(eb->i915->mm.userptr_wq);
+			mutex_lock(&eb->i915->drm.struct_mutex);
+			continue;
+		}
+
 		switch (pass++) {
 		case 0:
 			break;
@@ -689,13 +685,20 @@ static int eb_reserve(struct i915_execbuffer *eb)
 			err = i915_gem_evict_vm(eb->context->vm);
 			mutex_unlock(&eb->context->vm->mutex);
 			if (err)
-				return err;
+				goto unlock;
 			break;
 
 		default:
-			return -ENOSPC;
+			err = -ENOSPC;
+			goto unlock;
 		}
+
+		pin_flags = PIN_USER;
 	} while (1);
+
+unlock:
+	mutex_unlock(&eb->i915->drm.struct_mutex);
+	return err;
 }
 
 static unsigned int eb_batch_index(const struct i915_execbuffer *eb)
@@ -732,17 +735,14 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 	unsigned int i, batch;
 	int err;
 
+	if (unlikely(i915_gem_context_is_closed(eb->gem_context)))
+		return -ENOENT;
+
 	INIT_LIST_HEAD(&eb->relocs);
 	INIT_LIST_HEAD(&eb->unbound);
 
 	batch = eb_batch_index(eb);
 
-	mutex_lock(&eb->gem_context->mutex);
-	if (unlikely(i915_gem_context_is_closed(eb->gem_context))) {
-		err = -ENOENT;
-		goto err_ctx;
-	}
-
 	for (i = 0; i < eb->buffer_count; i++) {
 		u32 handle = eb->exec[i].handle;
 		struct i915_lut_handle *lut;
@@ -787,45 +787,37 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 		i915_gem_object_unlock(obj);
 
 add_vma:
-		err = eb_add_vma(eb, i, batch, vma);
+		err = eb_validate_vma(eb, &eb->exec[i], vma);
 		if (unlikely(err))
 			goto err_vma;
 
-		GEM_BUG_ON(vma != eb->vma[i]);
-		GEM_BUG_ON(vma->exec_flags != &eb->flags[i]);
-		GEM_BUG_ON(drm_mm_node_allocated(&vma->node) &&
-			   eb_vma_misplaced(&eb->exec[i], vma, eb->flags[i]));
+		eb_add_vma(eb, i, batch, vma);
 	}
 
-	mutex_unlock(&eb->gem_context->mutex);
-
-	eb->args->flags |= __EXEC_VALIDATED;
-	return eb_reserve(eb);
+	return 0;
 
 err_obj:
 	i915_gem_object_put(obj);
 err_vma:
-	eb->vma[i] = NULL;
-err_ctx:
-	mutex_unlock(&eb->gem_context->mutex);
+	eb->vma[i].vma = NULL;
 	return err;
 }
 
-static struct i915_vma *
+static struct eb_vma *
 eb_get_vma(const struct i915_execbuffer *eb, unsigned long handle)
 {
 	if (eb->lut_size < 0) {
 		if (handle >= -eb->lut_size)
 			return NULL;
-		return eb->vma[handle];
+		return &eb->vma[handle];
 	} else {
 		struct hlist_head *head;
-		struct i915_vma *vma;
+		struct eb_vma *ev;
 
 		head = &eb->buckets[hash_32(handle, eb->lut_size)];
-		hlist_for_each_entry(vma, head, exec_node) {
-			if (vma->exec_handle == handle)
-				return vma;
+		hlist_for_each_entry(ev, head, node) {
+			if (ev->handle == handle)
+				return ev;
 		}
 		return NULL;
 	}
@@ -837,32 +829,21 @@ static void eb_release_vmas(const struct i915_execbuffer *eb)
 	unsigned int i;
 
 	for (i = 0; i < count; i++) {
-		struct i915_vma *vma = eb->vma[i];
-		unsigned int flags = eb->flags[i];
+		struct eb_vma *ev = &eb->vma[i];
+		struct i915_vma *vma = ev->vma;
 
 		if (!vma)
 			break;
 
-		GEM_BUG_ON(vma->exec_flags != &eb->flags[i]);
-		vma->exec_flags = NULL;
-		eb->vma[i] = NULL;
+		eb->vma[i].vma = NULL;
 
-		if (flags & __EXEC_OBJECT_HAS_PIN)
-			__eb_unreserve_vma(vma, flags);
+		if (ev->flags & __EXEC_OBJECT_HAS_PIN)
+			__eb_unreserve_vma(vma, ev->flags);
 
-		if (flags & __EXEC_OBJECT_HAS_REF)
-			i915_vma_put(vma);
+		i915_vma_put(vma);
 	}
 }
 
-static void eb_reset_vmas(const struct i915_execbuffer *eb)
-{
-	eb_release_vmas(eb);
-	if (eb->lut_size > 0)
-		memset(eb->buckets, 0,
-		       sizeof(struct hlist_head) << eb->lut_size);
-}
-
 static void eb_destroy(const struct i915_execbuffer *eb)
 {
 	GEM_BUG_ON(eb->reloc_cache.rq);
@@ -1198,7 +1179,7 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb,
 	goto out_pool;
 
 skip_request:
-	i915_request_skip(rq, err);
+	i915_request_set_error_once(rq, err);
 err_request:
 	i915_request_add(rq);
 err_unpin:
@@ -1329,11 +1310,11 @@ out:
 
 static u64
 eb_relocate_entry(struct i915_execbuffer *eb,
-		  struct i915_vma *vma,
+		  struct eb_vma *ev,
 		  const struct drm_i915_gem_relocation_entry *reloc)
 {
 	struct drm_i915_private *i915 = eb->i915;
-	struct i915_vma *target;
+	struct eb_vma *target;
 	int err;
 
 	/* we've already hold a reference to all valid objects */
@@ -1365,7 +1346,7 @@ eb_relocate_entry(struct i915_execbuffer *eb,
 	}
 
 	if (reloc->write_domain) {
-		*target->exec_flags |= EXEC_OBJECT_WRITE;
+		target->flags |= EXEC_OBJECT_WRITE;
 
 		/*
 		 * Sandybridge PPGTT errata: We need a global gtt mapping
@@ -1375,7 +1356,8 @@ eb_relocate_entry(struct i915_execbuffer *eb,
 		 */
 		if (reloc->write_domain == I915_GEM_DOMAIN_INSTRUCTION &&
 		    IS_GEN(eb->i915, 6)) {
-			err = i915_vma_bind(target, target->obj->cache_level,
+			err = i915_vma_bind(target->vma,
+					    target->vma->obj->cache_level,
 					    PIN_GLOBAL, NULL);
 			if (WARN_ONCE(err,
 				      "Unexpected failure to bind target VMA!"))
@@ -1388,17 +1370,17 @@ eb_relocate_entry(struct i915_execbuffer *eb,
 	 * more work needs to be done.
 	 */
 	if (!DBG_FORCE_RELOC &&
-	    gen8_canonical_addr(target->node.start) == reloc->presumed_offset)
+	    gen8_canonical_addr(target->vma->node.start) == reloc->presumed_offset)
 		return 0;
 
 	/* Check that the relocation address is valid... */
 	if (unlikely(reloc->offset >
-		     vma->size - (eb->reloc_cache.use_64bit_reloc ? 8 : 4))) {
+		     ev->vma->size - (eb->reloc_cache.use_64bit_reloc ? 8 : 4))) {
 		drm_dbg(&i915->drm, "Relocation beyond object bounds: "
 			  "target %d offset %d size %d.\n",
 			  reloc->target_handle,
 			  (int)reloc->offset,
-			  (int)vma->size);
+			  (int)ev->vma->size);
 		return -EINVAL;
 	}
 	if (unlikely(reloc->offset & 3)) {
@@ -1417,18 +1399,18 @@ eb_relocate_entry(struct i915_execbuffer *eb,
 	 * do relocations we are already stalling, disable the user's opt
 	 * out of our synchronisation.
 	 */
-	*vma->exec_flags &= ~EXEC_OBJECT_ASYNC;
+	ev->flags &= ~EXEC_OBJECT_ASYNC;
 
 	/* and update the user's relocation entry */
-	return relocate_entry(vma, reloc, eb, target);
+	return relocate_entry(ev->vma, reloc, eb, target->vma);
 }
 
-static int eb_relocate_vma(struct i915_execbuffer *eb, struct i915_vma *vma)
+static int eb_relocate_vma(struct i915_execbuffer *eb, struct eb_vma *ev)
 {
 #define N_RELOC(x) ((x) / sizeof(struct drm_i915_gem_relocation_entry))
 	struct drm_i915_gem_relocation_entry stack[N_RELOC(512)];
 	struct drm_i915_gem_relocation_entry __user *urelocs;
-	const struct drm_i915_gem_exec_object2 *entry = exec_entry(eb, vma);
+	const struct drm_i915_gem_exec_object2 *entry = ev->exec;
 	unsigned int remain;
 
 	urelocs = u64_to_user_ptr(entry->relocs_ptr);
@@ -1458,9 +1440,7 @@ static int eb_relocate_vma(struct i915_execbuffer *eb, struct i915_vma *vma)
 		 * we would try to acquire the struct mutex again. Obviously
 		 * this is bad and so lockdep complains vehemently.
 		 */
-		pagefault_disable();
-		copied = __copy_from_user_inatomic(r, urelocs, count * sizeof(r[0]));
-		pagefault_enable();
+		copied = __copy_from_user(r, urelocs, count * sizeof(r[0]));
 		if (unlikely(copied)) {
 			remain = -EFAULT;
 			goto out;
@@ -1468,7 +1448,7 @@ static int eb_relocate_vma(struct i915_execbuffer *eb, struct i915_vma *vma)
 
 		remain -= count;
 		do {
-			u64 offset = eb_relocate_entry(eb, vma, r);
+			u64 offset = eb_relocate_entry(eb, ev, r);
 
 			if (likely(offset == 0)) {
 			} else if ((s64)offset < 0) {
@@ -1510,278 +1490,34 @@ out:
 	return remain;
 }
 
-static int
-eb_relocate_vma_slow(struct i915_execbuffer *eb, struct i915_vma *vma)
-{
-	const struct drm_i915_gem_exec_object2 *entry = exec_entry(eb, vma);
-	struct drm_i915_gem_relocation_entry *relocs =
-		u64_to_ptr(typeof(*relocs), entry->relocs_ptr);
-	unsigned int i;
-	int err;
-
-	for (i = 0; i < entry->relocation_count; i++) {
-		u64 offset = eb_relocate_entry(eb, vma, &relocs[i]);
-
-		if ((s64)offset < 0) {
-			err = (int)offset;
-			goto err;
-		}
-	}
-	err = 0;
-err:
-	reloc_cache_reset(&eb->reloc_cache);
-	return err;
-}
-
-static int check_relocations(const struct drm_i915_gem_exec_object2 *entry)
-{
-	const char __user *addr, *end;
-	unsigned long size;
-	char __maybe_unused c;
-
-	size = entry->relocation_count;
-	if (size == 0)
-		return 0;
-
-	if (size > N_RELOC(ULONG_MAX))
-		return -EINVAL;
-
-	addr = u64_to_user_ptr(entry->relocs_ptr);
-	size *= sizeof(struct drm_i915_gem_relocation_entry);
-	if (!access_ok(addr, size))
-		return -EFAULT;
-
-	end = addr + size;
-	for (; addr < end; addr += PAGE_SIZE) {
-		int err = __get_user(c, addr);
-		if (err)
-			return err;
-	}
-	return __get_user(c, end - 1);
-}
-
-static int eb_copy_relocations(const struct i915_execbuffer *eb)
+static int eb_relocate(struct i915_execbuffer *eb)
 {
-	struct drm_i915_gem_relocation_entry *relocs;
-	const unsigned int count = eb->buffer_count;
-	unsigned int i;
 	int err;
 
-	for (i = 0; i < count; i++) {
-		const unsigned int nreloc = eb->exec[i].relocation_count;
-		struct drm_i915_gem_relocation_entry __user *urelocs;
-		unsigned long size;
-		unsigned long copied;
-
-		if (nreloc == 0)
-			continue;
-
-		err = check_relocations(&eb->exec[i]);
-		if (err)
-			goto err;
-
-		urelocs = u64_to_user_ptr(eb->exec[i].relocs_ptr);
-		size = nreloc * sizeof(*relocs);
-
-		relocs = kvmalloc_array(size, 1, GFP_KERNEL);
-		if (!relocs) {
-			err = -ENOMEM;
-			goto err;
-		}
-
-		/* copy_from_user is limited to < 4GiB */
-		copied = 0;
-		do {
-			unsigned int len =
-				min_t(u64, BIT_ULL(31), size - copied);
-
-			if (__copy_from_user((char *)relocs + copied,
-					     (char __user *)urelocs + copied,
-					     len))
-				goto end;
-
-			copied += len;
-		} while (copied < size);
-
-		/*
-		 * As we do not update the known relocation offsets after
-		 * relocating (due to the complexities in lock handling),
-		 * we need to mark them as invalid now so that we force the
-		 * relocation processing next time. Just in case the target
-		 * object is evicted and then rebound into its old
-		 * presumed_offset before the next execbuffer - if that
-		 * happened we would make the mistake of assuming that the
-		 * relocations were valid.
-		 */
-		if (!user_access_begin(urelocs, size))
-			goto end;
-
-		for (copied = 0; copied < nreloc; copied++)
-			unsafe_put_user(-1,
-					&urelocs[copied].presumed_offset,
-					end_user);
-		user_access_end();
-
-		eb->exec[i].relocs_ptr = (uintptr_t)relocs;
-	}
-
-	return 0;
-
-end_user:
-	user_access_end();
-end:
-	kvfree(relocs);
-	err = -EFAULT;
-err:
-	while (i--) {
-		relocs = u64_to_ptr(typeof(*relocs), eb->exec[i].relocs_ptr);
-		if (eb->exec[i].relocation_count)
-			kvfree(relocs);
-	}
-	return err;
-}
-
-static int eb_prefault_relocations(const struct i915_execbuffer *eb)
-{
-	const unsigned int count = eb->buffer_count;
-	unsigned int i;
-
-	for (i = 0; i < count; i++) {
-		int err;
-
-		err = check_relocations(&eb->exec[i]);
-		if (err)
-			return err;
-	}
-
-	return 0;
-}
-
-static noinline int eb_relocate_slow(struct i915_execbuffer *eb)
-{
-	struct drm_device *dev = &eb->i915->drm;
-	bool have_copy = false;
-	struct i915_vma *vma;
-	int err = 0;
-
-repeat:
-	if (signal_pending(current)) {
-		err = -ERESTARTSYS;
-		goto out;
-	}
-
-	/* We may process another execbuffer during the unlock... */
-	eb_reset_vmas(eb);
-	mutex_unlock(&dev->struct_mutex);
-
-	/*
-	 * We take 3 passes through the slowpatch.
-	 *
-	 * 1 - we try to just prefault all the user relocation entries and
-	 * then attempt to reuse the atomic pagefault disabled fast path again.
-	 *
-	 * 2 - we copy the user entries to a local buffer here outside of the
-	 * local and allow ourselves to wait upon any rendering before
-	 * relocations
-	 *
-	 * 3 - we already have a local copy of the relocation entries, but
-	 * were interrupted (EAGAIN) whilst waiting for the objects, try again.
-	 */
-	if (!err) {
-		err = eb_prefault_relocations(eb);
-	} else if (!have_copy) {
-		err = eb_copy_relocations(eb);
-		have_copy = err == 0;
-	} else {
-		cond_resched();
-		err = 0;
-	}
-	if (err) {
-		mutex_lock(&dev->struct_mutex);
-		goto out;
-	}
-
-	/* A frequent cause for EAGAIN are currently unavailable client pages */
-	flush_workqueue(eb->i915->mm.userptr_wq);
-
-	err = i915_mutex_lock_interruptible(dev);
-	if (err) {
-		mutex_lock(&dev->struct_mutex);
-		goto out;
-	}
-
-	/* reacquire the objects */
+	mutex_lock(&eb->gem_context->mutex);
 	err = eb_lookup_vmas(eb);
+	mutex_unlock(&eb->gem_context->mutex);
 	if (err)
-		goto err;
-
-	GEM_BUG_ON(!eb->batch);
-
-	list_for_each_entry(vma, &eb->relocs, reloc_link) {
-		if (!have_copy) {
-			pagefault_disable();
-			err = eb_relocate_vma(eb, vma);
-			pagefault_enable();
-			if (err)
-				goto repeat;
-		} else {
-			err = eb_relocate_vma_slow(eb, vma);
-			if (err)
-				goto err;
-		}
-	}
-
-	/*
-	 * Leave the user relocations as are, this is the painfully slow path,
-	 * and we want to avoid the complication of dropping the lock whilst
-	 * having buffers reserved in the aperture and so causing spurious
-	 * ENOSPC for random operations.
-	 */
-
-err:
-	if (err == -EAGAIN)
-		goto repeat;
-
-out:
-	if (have_copy) {
-		const unsigned int count = eb->buffer_count;
-		unsigned int i;
-
-		for (i = 0; i < count; i++) {
-			const struct drm_i915_gem_exec_object2 *entry =
-				&eb->exec[i];
-			struct drm_i915_gem_relocation_entry *relocs;
-
-			if (!entry->relocation_count)
-				continue;
+		return err;
 
-			relocs = u64_to_ptr(typeof(*relocs), entry->relocs_ptr);
-			kvfree(relocs);
-		}
+	if (!list_empty(&eb->unbound)) {
+		err = eb_reserve(eb);
+		if (err)
+			return err;
 	}
 
-	return err;
-}
-
-static int eb_relocate(struct i915_execbuffer *eb)
-{
-	if (eb_lookup_vmas(eb))
-		goto slow;
-
 	/* The objects are in their final locations, apply the relocations. */
 	if (eb->args->flags & __EXEC_HAS_RELOC) {
-		struct i915_vma *vma;
+		struct eb_vma *ev;
 
-		list_for_each_entry(vma, &eb->relocs, reloc_link) {
-			if (eb_relocate_vma(eb, vma))
-				goto slow;
+		list_for_each_entry(ev, &eb->relocs, reloc_link) {
+			err = eb_relocate_vma(eb, ev);
+			if (err)
+				return err;
 		}
 	}
 
 	return 0;
-
-slow:
-	return eb_relocate_slow(eb);
 }
 
 static int eb_move_to_gpu(struct i915_execbuffer *eb)
@@ -1794,27 +1530,19 @@ static int eb_move_to_gpu(struct i915_execbuffer *eb)
 	ww_acquire_init(&acquire, &reservation_ww_class);
 
 	for (i = 0; i < count; i++) {
-		struct i915_vma *vma = eb->vma[i];
+		struct eb_vma *ev = &eb->vma[i];
+		struct i915_vma *vma = ev->vma;
 
 		err = ww_mutex_lock_interruptible(&vma->resv->lock, &acquire);
-		if (!err)
-			continue;
-
-		GEM_BUG_ON(err == -EALREADY); /* No duplicate vma */
-
 		if (err == -EDEADLK) {
 			GEM_BUG_ON(i == 0);
 			do {
 				int j = i - 1;
 
-				ww_mutex_unlock(&eb->vma[j]->resv->lock);
+				ww_mutex_unlock(&eb->vma[j].vma->resv->lock);
 
-				swap(eb->flags[i], eb->flags[j]);
 				swap(eb->vma[i],  eb->vma[j]);
-				eb->vma[i]->exec_flags = &eb->flags[i];
 			} while (--i);
-			GEM_BUG_ON(vma != eb->vma[0]);
-			vma->exec_flags = &eb->flags[0];
 
 			err = ww_mutex_lock_slow_interruptible(&vma->resv->lock,
 							       &acquire);
@@ -1825,8 +1553,9 @@ static int eb_move_to_gpu(struct i915_execbuffer *eb)
 	ww_acquire_done(&acquire);
 
 	while (i--) {
-		unsigned int flags = eb->flags[i];
-		struct i915_vma *vma = eb->vma[i];
+		struct eb_vma *ev = &eb->vma[i];
+		struct i915_vma *vma = ev->vma;
+		unsigned int flags = ev->flags;
 		struct drm_i915_gem_object *obj = vma->obj;
 
 		assert_vma_held(vma);
@@ -1870,10 +1599,9 @@ static int eb_move_to_gpu(struct i915_execbuffer *eb)
 		i915_vma_unlock(vma);
 
 		__eb_unreserve_vma(vma, flags);
-		vma->exec_flags = NULL;
+		i915_vma_put(vma);
 
-		if (unlikely(flags & __EXEC_OBJECT_HAS_REF))
-			i915_vma_put(vma);
+		ev->vma = NULL;
 	}
 	ww_acquire_fini(&acquire);
 
@@ -1887,7 +1615,7 @@ static int eb_move_to_gpu(struct i915_execbuffer *eb)
 	return 0;
 
 err_skip:
-	i915_request_skip(eb->request, err);
+	i915_request_set_error_once(eb->request, err);
 	return err;
 }
 
@@ -2008,7 +1736,7 @@ static int eb_parse_pipeline(struct i915_execbuffer *eb,
 	if (!pw)
 		return -ENOMEM;
 
-	err = i915_active_acquire(&eb->batch->active);
+	err = i915_active_acquire(&eb->batch->vma->active);
 	if (err)
 		goto err_free;
 
@@ -2025,7 +1753,7 @@ static int eb_parse_pipeline(struct i915_execbuffer *eb,
 	dma_fence_work_init(&pw->base, &eb_parse_ops);
 
 	pw->engine = eb->engine;
-	pw->batch = eb->batch;
+	pw->batch = eb->batch->vma;
 	pw->batch_offset = eb->batch_start_offset;
 	pw->batch_length = eb->batch_len;
 	pw->shadow = shadow;
@@ -2067,7 +1795,7 @@ err_trampoline:
 err_shadow:
 	i915_active_release(&shadow->active);
 err_batch:
-	i915_active_release(&eb->batch->active);
+	i915_active_release(&eb->batch->vma->active);
 err_free:
 	kfree(pw);
 	return err;
@@ -2130,15 +1858,12 @@ static int eb_parse(struct i915_execbuffer *eb)
 	if (err)
 		goto err_trampoline;
 
-	eb->vma[eb->buffer_count] = i915_vma_get(shadow);
-	eb->flags[eb->buffer_count] =
-		__EXEC_OBJECT_HAS_PIN | __EXEC_OBJECT_HAS_REF;
-	shadow->exec_flags = &eb->flags[eb->buffer_count];
-	eb->buffer_count++;
+	eb->vma[eb->buffer_count].vma = i915_vma_get(shadow);
+	eb->vma[eb->buffer_count].flags = __EXEC_OBJECT_HAS_PIN;
+	eb->batch = &eb->vma[eb->buffer_count++];
 
 	eb->trampoline = trampoline;
 	eb->batch_start_offset = 0;
-	eb->batch = shadow;
 
 	shadow->private = pool;
 	return 0;
@@ -2165,7 +1890,7 @@ add_to_client(struct i915_request *rq, struct drm_file *file)
 	spin_unlock(&file_priv->mm.lock);
 }
 
-static int eb_submit(struct i915_execbuffer *eb)
+static int eb_submit(struct i915_execbuffer *eb, struct i915_vma *batch)
 {
 	int err;
 
@@ -2192,7 +1917,7 @@ static int eb_submit(struct i915_execbuffer *eb)
 	}
 
 	err = eb->engine->emit_bb_start(eb->request,
-					eb->batch->node.start +
+					batch->node.start +
 					eb->batch_start_offset,
 					eb->batch_len,
 					eb->batch_flags);
@@ -2327,15 +2052,22 @@ static int __eb_pin_engine(struct i915_execbuffer *eb, struct intel_context *ce)
 	intel_context_timeline_unlock(tl);
 
 	if (rq) {
-		if (i915_request_wait(rq,
-				      I915_WAIT_INTERRUPTIBLE,
-				      MAX_SCHEDULE_TIMEOUT) < 0) {
-			i915_request_put(rq);
-			err = -EINTR;
-			goto err_exit;
-		}
+		bool nonblock = eb->file->filp->f_flags & O_NONBLOCK;
+		long timeout;
 
+		timeout = MAX_SCHEDULE_TIMEOUT;
+		if (nonblock)
+			timeout = 0;
+
+		timeout = i915_request_wait(rq,
+					    I915_WAIT_INTERRUPTIBLE,
+					    timeout);
 		i915_request_put(rq);
+
+		if (timeout < 0) {
+			err = nonblock ? -EWOULDBLOCK : timeout;
+			goto err_exit;
+		}
 	}
 
 	eb->engine = ce->engine;
@@ -2560,6 +2292,73 @@ signal_fence_array(struct i915_execbuffer *eb,
 	}
 }
 
+static void retire_requests(struct intel_timeline *tl, struct i915_request *end)
+{
+	struct i915_request *rq, *rn;
+
+	list_for_each_entry_safe(rq, rn, &tl->requests, link)
+		if (rq == end || !i915_request_retire(rq))
+			break;
+}
+
+static void eb_request_add(struct i915_execbuffer *eb)
+{
+	struct i915_request *rq = eb->request;
+	struct intel_timeline * const tl = i915_request_timeline(rq);
+	struct i915_sched_attr attr = {};
+	struct i915_request *prev;
+
+	lockdep_assert_held(&tl->mutex);
+	lockdep_unpin_lock(&tl->mutex, rq->cookie);
+
+	trace_i915_request_add(rq);
+
+	prev = __i915_request_commit(rq);
+
+	/* Check that the context wasn't destroyed before submission */
+	if (likely(rcu_access_pointer(eb->context->gem_context))) {
+		attr = eb->gem_context->sched;
+
+		/*
+		 * Boost actual workloads past semaphores!
+		 *
+		 * With semaphores we spin on one engine waiting for another,
+		 * simply to reduce the latency of starting our work when
+		 * the signaler completes. However, if there is any other
+		 * work that we could be doing on this engine instead, that
+		 * is better utilisation and will reduce the overall duration
+		 * of the current work. To avoid PI boosting a semaphore
+		 * far in the distance past over useful work, we keep a history
+		 * of any semaphore use along our dependency chain.
+		 */
+		if (!(rq->sched.flags & I915_SCHED_HAS_SEMAPHORE_CHAIN))
+			attr.priority |= I915_PRIORITY_NOSEMAPHORE;
+
+		/*
+		 * Boost priorities to new clients (new request flows).
+		 *
+		 * Allow interactive/synchronous clients to jump ahead of
+		 * the bulk clients. (FQ_CODEL)
+		 */
+		if (list_empty(&rq->sched.signalers_list))
+			attr.priority |= I915_PRIORITY_WAIT;
+	} else {
+		/* Serialise with context_close via the add_to_timeline */
+		i915_request_set_error_once(rq, -ENOENT);
+		__i915_request_skip(rq);
+	}
+
+	local_bh_disable();
+	__i915_request_queue(rq, &attr);
+	local_bh_enable(); /* Kick the execlists tasklet if just scheduled */
+
+	/* Try to clean up the client's timeline after submitting the request */
+	if (prev)
+		retire_requests(tl, prev);
+
+	mutex_unlock(&tl->mutex);
+}
+
 static int
 i915_gem_do_execbuffer(struct drm_device *dev,
 		       struct drm_file *file,
@@ -2572,6 +2371,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	struct dma_fence *in_fence = NULL;
 	struct dma_fence *exec_fence = NULL;
 	struct sync_file *out_fence = NULL;
+	struct i915_vma *batch;
 	int out_fence_fd = -1;
 	int err;
 
@@ -2586,9 +2386,8 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 		args->flags |= __EXEC_HAS_RELOC;
 
 	eb.exec = exec;
-	eb.vma = (struct i915_vma **)(exec + args->buffer_count + 1);
-	eb.vma[0] = NULL;
-	eb.flags = (unsigned int *)(eb.vma + args->buffer_count + 1);
+	eb.vma = (struct eb_vma *)(exec + args->buffer_count + 1);
+	eb.vma[0].vma = NULL;
 
 	eb.invalid_flags = __EXEC_OBJECT_UNKNOWN_FLAGS;
 	reloc_cache_init(&eb.reloc_cache, eb.i915);
@@ -2656,10 +2455,6 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	if (unlikely(err))
 		goto err_context;
 
-	err = i915_mutex_lock_interruptible(dev);
-	if (err)
-		goto err_engine;
-
 	err = eb_relocate(&eb);
 	if (err) {
 		/*
@@ -2673,21 +2468,23 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 		goto err_vma;
 	}
 
-	if (unlikely(*eb.batch->exec_flags & EXEC_OBJECT_WRITE)) {
+	if (unlikely(eb.batch->flags & EXEC_OBJECT_WRITE)) {
 		drm_dbg(&i915->drm,
 			"Attempting to use self-modifying batch buffer\n");
 		err = -EINVAL;
 		goto err_vma;
 	}
-	if (eb.batch_start_offset > eb.batch->size ||
-	    eb.batch_len > eb.batch->size - eb.batch_start_offset) {
+
+	if (range_overflows_t(u64,
+			      eb.batch_start_offset, eb.batch_len,
+			      eb.batch->vma->size)) {
 		drm_dbg(&i915->drm, "Attempting to use out-of-bounds batch\n");
 		err = -EINVAL;
 		goto err_vma;
 	}
 
 	if (eb.batch_len == 0)
-		eb.batch_len = eb.batch->size - eb.batch_start_offset;
+		eb.batch_len = eb.batch->vma->size - eb.batch_start_offset;
 
 	err = eb_parse(&eb);
 	if (err)
@@ -2697,6 +2494,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	 * snb/ivb/vlv conflate the "batch in ppgtt" bit with the "non-secure
 	 * batch" bit. Hence we need to pin secure batches into the global gtt.
 	 * hsw should have this fixed, but bdw mucks it up again. */
+	batch = eb.batch->vma;
 	if (eb.batch_flags & I915_DISPATCH_SECURE) {
 		struct i915_vma *vma;
 
@@ -2710,13 +2508,13 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 		 *   fitting due to fragmentation.
 		 * So this is actually safe.
 		 */
-		vma = i915_gem_object_ggtt_pin(eb.batch->obj, NULL, 0, 0, 0);
+		vma = i915_gem_object_ggtt_pin(batch->obj, NULL, 0, 0, 0);
 		if (IS_ERR(vma)) {
 			err = PTR_ERR(vma);
 			goto err_parse;
 		}
 
-		eb.batch = vma;
+		batch = vma;
 	}
 
 	/* All GPU relocation batches must be submitted prior to the user rq */
@@ -2763,16 +2561,16 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	 * inactive_list and lose its active reference. Hence we do not need
 	 * to explicitly hold another reference here.
 	 */
-	eb.request->batch = eb.batch;
-	if (eb.batch->private)
-		intel_engine_pool_mark_active(eb.batch->private, eb.request);
+	eb.request->batch = batch;
+	if (batch->private)
+		intel_engine_pool_mark_active(batch->private, eb.request);
 
 	trace_i915_request_queue(eb.request, eb.batch_flags);
-	err = eb_submit(&eb);
+	err = eb_submit(&eb, batch);
 err_request:
 	add_to_client(eb.request, file);
 	i915_request_get(eb.request);
-	i915_request_add(eb.request);
+	eb_request_add(&eb);
 
 	if (fences)
 		signal_fence_array(&eb, fences);
@@ -2791,17 +2589,15 @@ err_request:
 
 err_batch_unpin:
 	if (eb.batch_flags & I915_DISPATCH_SECURE)
-		i915_vma_unpin(eb.batch);
+		i915_vma_unpin(batch);
 err_parse:
-	if (eb.batch->private)
-		intel_engine_pool_put(eb.batch->private);
+	if (batch->private)
+		intel_engine_pool_put(batch->private);
 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);
 err_context:
 	i915_gem_context_put(eb.gem_context);
@@ -2819,9 +2615,7 @@ err_in_fence:
 
 static size_t eb_element_size(void)
 {
-	return (sizeof(struct drm_i915_gem_exec_object2) +
-		sizeof(struct i915_vma *) +
-		sizeof(unsigned int));
+	return sizeof(struct drm_i915_gem_exec_object2) + sizeof(struct eb_vma);
 }
 
 static bool check_buffer_count(size_t count)
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_internal.c b/drivers/gpu/drm/i915/gem/i915_gem_internal.c
index 9cfb0e41ff06..cbbff81aa0af 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_internal.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_internal.c
@@ -8,8 +8,6 @@
 #include <linux/slab.h>
 #include <linux/swiotlb.h>
 
-#include <drm/i915_drm.h>
-
 #include "i915_drv.h"
 #include "i915_gem.h"
 #include "i915_gem_object.h"
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_mman.c b/drivers/gpu/drm/i915/gem/i915_gem_mman.c
index e8cccc131c40..b39c24dae64e 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_mman.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_mman.c
@@ -775,7 +775,7 @@ static struct file *mmap_singleton(struct drm_i915_private *i915)
 	struct file *file;
 
 	rcu_read_lock();
-	file = i915->gem.mmap_singleton;
+	file = READ_ONCE(i915->gem.mmap_singleton);
 	if (file && !get_file_rcu(file))
 		file = NULL;
 	rcu_read_unlock();
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h
index e44a2f40b520..2faa481cc18f 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h
@@ -11,8 +11,6 @@
 #include <drm/drm_file.h>
 #include <drm/drm_device.h>
 
-#include <drm/i915_drm.h>
-
 #include "display/intel_frontbuffer.h"
 #include "i915_gem_object_types.h"
 #include "i915_gem_gtt.h"
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c
index 70809d8897cd..e00792158f13 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c
@@ -186,7 +186,7 @@ int i915_gem_object_fill_blt(struct drm_i915_gem_object *obj,
 					0);
 out_request:
 	if (unlikely(err))
-		i915_request_skip(rq, err);
+		i915_request_set_error_once(rq, err);
 
 	i915_request_add(rq);
 out_batch:
@@ -196,6 +196,17 @@ out_unpin:
 	return err;
 }
 
+/* Wa_1209644611:icl,ehl */
+static bool wa_1209644611_applies(struct drm_i915_private *i915, u32 size)
+{
+	u32 height = size >> PAGE_SHIFT;
+
+	if (!IS_GEN(i915, 11))
+		return false;
+
+	return height % 4 == 3 && height <= 8;
+}
+
 struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce,
 					 struct i915_vma *src,
 					 struct i915_vma *dst)
@@ -237,7 +248,8 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce,
 		size = min_t(u64, rem, block_size);
 		GEM_BUG_ON(size >> PAGE_SHIFT > S16_MAX);
 
-		if (INTEL_GEN(i915) >= 9) {
+		if (INTEL_GEN(i915) >= 9 &&
+		    !wa_1209644611_applies(i915, size)) {
 			*cmd++ = GEN9_XY_FAST_COPY_BLT_CMD | (10 - 2);
 			*cmd++ = BLT_DEPTH_32 | PAGE_SIZE;
 			*cmd++ = 0;
@@ -385,7 +397,7 @@ out_unlock:
 	drm_gem_unlock_reservations(objs, ARRAY_SIZE(objs), &acquire);
 out_request:
 	if (unlikely(err))
-		i915_request_skip(rq, err);
+		i915_request_set_error_once(rq, err);
 
 	i915_request_add(rq);
 out_batch:
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_phys.c b/drivers/gpu/drm/i915/gem/i915_gem_phys.c
index b07bb40edd5a..698e22420dc5 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_phys.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_phys.c
@@ -194,10 +194,11 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align)
 	/* Perma-pin (until release) the physical set of pages */
 	__i915_gem_object_pin_pages(obj);
 
-	if (!IS_ERR_OR_NULL(pages)) {
+	if (!IS_ERR_OR_NULL(pages))
 		i915_gem_shmem_ops.put_pages(obj, pages);
-		i915_gem_object_release_memory_region(obj);
-	}
+
+	i915_gem_object_release_memory_region(obj);
+
 	mutex_unlock(&obj->mm.lock);
 	return 0;
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
index 830d3f96e1f6..03e5eb4c99d1 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
@@ -12,7 +12,6 @@
 #include <linux/pci.h>
 #include <linux/dma-buf.h>
 #include <linux/vmalloc.h>
-#include <drm/i915_drm.h>
 
 #include "i915_trace.h"
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_stolen.c b/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
index 491cfbaaa330..5557dfa83a7b 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
@@ -13,6 +13,7 @@
 #include "gem/i915_gem_region.h"
 #include "i915_drv.h"
 #include "i915_gem_stolen.h"
+#include "i915_vgpu.h"
 
 /*
  * The BIOS typically reserves some of the system's memory for the exclusive
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_tiling.c b/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
index 6c7825a2dc2a..37f77aee1212 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
@@ -6,7 +6,6 @@
 
 #include <linux/string.h>
 #include <linux/bitops.h>
-#include <drm/i915_drm.h>
 
 #include "i915_drv.h"
 #include "i915_gem.h"
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_userptr.c b/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
index 63ead7a2b64a..7ffd7afeb7a5 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
@@ -10,8 +10,6 @@
 #include <linux/swap.h>
 #include <linux/sched/mm.h>
 
-#include <drm/i915_drm.h>
-
 #include "i915_drv.h"
 #include "i915_gem_ioctls.h"
 #include "i915_gem_object.h"
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 375d864736f3..54b86cf7f5d2 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
@@ -1004,7 +1004,7 @@ emit_rpcs_query(struct drm_i915_gem_object *obj,
 	return 0;
 
 skip_request:
-	i915_request_skip(rq, err);
+	i915_request_set_error_once(rq, err);
 err_request:
 	i915_request_add(rq);
 err_batch:
@@ -1559,7 +1559,7 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 
 	goto out_vm;
 skip_request:
-	i915_request_skip(rq, err);
+	i915_request_set_error_once(rq, err);
 err_request:
 	i915_request_add(rq);
 err_unpin:
@@ -1708,7 +1708,7 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 
 	goto out_vm;
 skip_request:
-	i915_request_skip(rq, err);
+	i915_request_set_error_once(rq, err);
 err_request:
 	i915_request_add(rq);
 err_unpin:
@@ -1809,7 +1809,6 @@ static int igt_vm_isolation(void *arg)
 
 	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;
@@ -1828,10 +1827,10 @@ static int igt_vm_isolation(void *arg)
 			u32 value = 0xc5c5c5c5;
 			u64 offset;
 
-			div64_u64_rem(i915_prandom_u64_state(&prng),
-				      vm_total, &offset);
-			offset = round_down(offset, alignof_dword);
-			offset += I915_GTT_PAGE_SIZE;
+			/* Leave enough space at offset 0 for the batch */
+			offset = igt_random_offset(&prng,
+						   I915_GTT_PAGE_SIZE, vm_total,
+						   sizeof(u32), alignof_dword);
 
 			err = write_to_scratch(ctx_a, engine,
 					       offset, 0xdeadbeef);
diff --git a/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c b/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c
index 6718da20f35d..772d8cba7da9 100644
--- a/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c
+++ b/drivers/gpu/drm/i915/gem/selftests/igt_gem_utils.c
@@ -159,7 +159,7 @@ int igt_gpu_fill_dw(struct intel_context *ce,
 	return 0;
 
 skip_request:
-	i915_request_skip(rq, err);
+	i915_request_set_error_once(rq, err);
 err_request:
 	i915_request_add(rq);
 err_batch:
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_context.c b/drivers/gpu/drm/i915/gem/selftests/mock_context.c
index b12ea1daa29d..e7e3c620f542 100644
--- a/drivers/gpu/drm/i915/gem/selftests/mock_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/mock_context.c
@@ -23,6 +23,9 @@ mock_context(struct drm_i915_private *i915,
 	INIT_LIST_HEAD(&ctx->link);
 	ctx->i915 = i915;
 
+	spin_lock_init(&ctx->stale.lock);
+	INIT_LIST_HEAD(&ctx->stale.engines);
+
 	i915_gem_context_set_persistence(ctx);
 
 	mutex_init(&ctx->engines_mutex);
diff --git a/drivers/gpu/drm/i915/gt/gen7_renderclear.c b/drivers/gpu/drm/i915/gt/gen7_renderclear.c
new file mode 100644
index 000000000000..de595b66a746
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/gen7_renderclear.c
@@ -0,0 +1,402 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "gen7_renderclear.h"
+#include "i915_drv.h"
+#include "intel_gpu_commands.h"
+
+#define MAX_URB_ENTRIES 64
+#define STATE_SIZE (4 * 1024)
+#define GT3_INLINE_DATA_DELAYS 0x1E00
+#define batch_advance(Y, CS) GEM_BUG_ON((Y)->end != (CS))
+
+struct cb_kernel {
+	const void *data;
+	u32 size;
+};
+
+#define CB_KERNEL(name) { .data = (name), .size = sizeof(name) }
+
+#include "ivb_clear_kernel.c"
+static const struct cb_kernel cb_kernel_ivb = CB_KERNEL(ivb_clear_kernel);
+
+#include "hsw_clear_kernel.c"
+static const struct cb_kernel cb_kernel_hsw = CB_KERNEL(hsw_clear_kernel);
+
+struct batch_chunk {
+	struct i915_vma *vma;
+	u32 offset;
+	u32 *start;
+	u32 *end;
+	u32 max_items;
+};
+
+struct batch_vals {
+	u32 max_primitives;
+	u32 max_urb_entries;
+	u32 cmd_size;
+	u32 state_size;
+	u32 state_start;
+	u32 batch_size;
+	u32 surface_height;
+	u32 surface_width;
+	u32 scratch_size;
+	u32 max_size;
+};
+
+static void
+batch_get_defaults(struct drm_i915_private *i915, struct batch_vals *bv)
+{
+	if (IS_HASWELL(i915)) {
+		bv->max_primitives = 280;
+		bv->max_urb_entries = MAX_URB_ENTRIES;
+		bv->surface_height = 16 * 16;
+		bv->surface_width = 32 * 2 * 16;
+	} else {
+		bv->max_primitives = 128;
+		bv->max_urb_entries = MAX_URB_ENTRIES / 2;
+		bv->surface_height = 16 * 8;
+		bv->surface_width = 32 * 16;
+	}
+	bv->cmd_size = bv->max_primitives * 4096;
+	bv->state_size = STATE_SIZE;
+	bv->state_start = bv->cmd_size;
+	bv->batch_size = bv->cmd_size + bv->state_size;
+	bv->scratch_size = bv->surface_height * bv->surface_width;
+	bv->max_size = bv->batch_size + bv->scratch_size;
+}
+
+static void batch_init(struct batch_chunk *bc,
+		       struct i915_vma *vma,
+		       u32 *start, u32 offset, u32 max_bytes)
+{
+	bc->vma = vma;
+	bc->offset = offset;
+	bc->start = start + bc->offset / sizeof(*bc->start);
+	bc->end = bc->start;
+	bc->max_items = max_bytes / sizeof(*bc->start);
+}
+
+static u32 batch_offset(const struct batch_chunk *bc, u32 *cs)
+{
+	return (cs - bc->start) * sizeof(*bc->start) + bc->offset;
+}
+
+static u32 batch_addr(const struct batch_chunk *bc)
+{
+	return bc->vma->node.start;
+}
+
+static void batch_add(struct batch_chunk *bc, const u32 d)
+{
+	GEM_BUG_ON((bc->end - bc->start) >= bc->max_items);
+	*bc->end++ = d;
+}
+
+static u32 *batch_alloc_items(struct batch_chunk *bc, u32 align, u32 items)
+{
+	u32 *map;
+
+	if (align) {
+		u32 *end = PTR_ALIGN(bc->end, align);
+
+		memset32(bc->end, 0, end - bc->end);
+		bc->end = end;
+	}
+
+	map = bc->end;
+	bc->end += items;
+
+	return map;
+}
+
+static u32 *batch_alloc_bytes(struct batch_chunk *bc, u32 align, u32 bytes)
+{
+	GEM_BUG_ON(!IS_ALIGNED(bytes, sizeof(*bc->start)));
+	return batch_alloc_items(bc, align, bytes / sizeof(*bc->start));
+}
+
+static u32
+gen7_fill_surface_state(struct batch_chunk *state,
+			const u32 dst_offset,
+			const struct batch_vals *bv)
+{
+	u32 surface_h = bv->surface_height;
+	u32 surface_w = bv->surface_width;
+	u32 *cs = batch_alloc_items(state, 32, 8);
+	u32 offset = batch_offset(state, cs);
+
+#define SURFACE_2D 1
+#define SURFACEFORMAT_B8G8R8A8_UNORM 0x0C0
+#define RENDER_CACHE_READ_WRITE 1
+
+	*cs++ = SURFACE_2D << 29 |
+		(SURFACEFORMAT_B8G8R8A8_UNORM << 18) |
+		(RENDER_CACHE_READ_WRITE << 8);
+
+	*cs++ = batch_addr(state) + dst_offset;
+
+	*cs++ = ((surface_h / 4 - 1) << 16) | (surface_w / 4 - 1);
+	*cs++ = surface_w;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+#define SHADER_CHANNELS(r, g, b, a) \
+	(((r) << 25) | ((g) << 22) | ((b) << 19) | ((a) << 16))
+	*cs++ = SHADER_CHANNELS(4, 5, 6, 7);
+	batch_advance(state, cs);
+
+	return offset;
+}
+
+static u32
+gen7_fill_binding_table(struct batch_chunk *state,
+			const struct batch_vals *bv)
+{
+	u32 surface_start = gen7_fill_surface_state(state, bv->batch_size, bv);
+	u32 *cs = batch_alloc_items(state, 32, 8);
+	u32 offset = batch_offset(state, cs);
+
+	*cs++ = surface_start - state->offset;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	batch_advance(state, cs);
+
+	return offset;
+}
+
+static u32
+gen7_fill_kernel_data(struct batch_chunk *state,
+		      const u32 *data,
+		      const u32 size)
+{
+	return batch_offset(state,
+			    memcpy(batch_alloc_bytes(state, 64, size),
+				   data, size));
+}
+
+static u32
+gen7_fill_interface_descriptor(struct batch_chunk *state,
+			       const struct batch_vals *bv,
+			       const struct cb_kernel *kernel,
+			       unsigned int count)
+{
+	u32 kernel_offset =
+		gen7_fill_kernel_data(state, kernel->data, kernel->size);
+	u32 binding_table = gen7_fill_binding_table(state, bv);
+	u32 *cs = batch_alloc_items(state, 32, 8 * count);
+	u32 offset = batch_offset(state, cs);
+
+	*cs++ = kernel_offset;
+	*cs++ = (1 << 7) | (1 << 13);
+	*cs++ = 0;
+	*cs++ = (binding_table - state->offset) | 1;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+
+	/* 1 - 63dummy idds */
+	memset32(cs, 0x00, (count - 1) * 8);
+	batch_advance(state, cs + (count - 1) * 8);
+
+	return offset;
+}
+
+static void
+gen7_emit_state_base_address(struct batch_chunk *batch,
+			     u32 surface_state_base)
+{
+	u32 *cs = batch_alloc_items(batch, 0, 12);
+
+	*cs++ = STATE_BASE_ADDRESS | (12 - 2);
+	/* general */
+	*cs++ = batch_addr(batch) | BASE_ADDRESS_MODIFY;
+	/* surface */
+	*cs++ = batch_addr(batch) | surface_state_base | BASE_ADDRESS_MODIFY;
+	/* dynamic */
+	*cs++ = batch_addr(batch) | BASE_ADDRESS_MODIFY;
+	/* indirect */
+	*cs++ = batch_addr(batch) | BASE_ADDRESS_MODIFY;
+	/* instruction */
+	*cs++ = batch_addr(batch) | BASE_ADDRESS_MODIFY;
+
+	/* general/dynamic/indirect/instruction access Bound */
+	*cs++ = 0;
+	*cs++ = BASE_ADDRESS_MODIFY;
+	*cs++ = 0;
+	*cs++ = BASE_ADDRESS_MODIFY;
+	*cs++ = 0;
+	*cs++ = 0;
+	batch_advance(batch, cs);
+}
+
+static void
+gen7_emit_vfe_state(struct batch_chunk *batch,
+		    const struct batch_vals *bv,
+		    u32 urb_size, u32 curbe_size,
+		    u32 mode)
+{
+	u32 urb_entries = bv->max_urb_entries;
+	u32 threads = bv->max_primitives - 1;
+	u32 *cs = batch_alloc_items(batch, 32, 8);
+
+	*cs++ = MEDIA_VFE_STATE | (8 - 2);
+
+	/* scratch buffer */
+	*cs++ = 0;
+
+	/* number of threads & urb entries for GPGPU vs Media Mode */
+	*cs++ = threads << 16 | urb_entries << 8 | mode << 2;
+
+	*cs++ = 0;
+
+	/* urb entry size & curbe size in 256 bits unit */
+	*cs++ = urb_size << 16 | curbe_size;
+
+	/* scoreboard */
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	batch_advance(batch, cs);
+}
+
+static void
+gen7_emit_interface_descriptor_load(struct batch_chunk *batch,
+				    const u32 interface_descriptor,
+				    unsigned int count)
+{
+	u32 *cs = batch_alloc_items(batch, 8, 4);
+
+	*cs++ = MEDIA_INTERFACE_DESCRIPTOR_LOAD | (4 - 2);
+	*cs++ = 0;
+	*cs++ = count * 8 * sizeof(*cs);
+
+	/*
+	 * interface descriptor address - it is relative to the dynamics base
+	 * address
+	 */
+	*cs++ = interface_descriptor;
+	batch_advance(batch, cs);
+}
+
+static void
+gen7_emit_media_object(struct batch_chunk *batch,
+		       unsigned int media_object_index)
+{
+	unsigned int x_offset = (media_object_index % 16) * 64;
+	unsigned int y_offset = (media_object_index / 16) * 16;
+	unsigned int inline_data_size;
+	unsigned int media_batch_size;
+	unsigned int i;
+	u32 *cs;
+
+	inline_data_size = 112 * 8;
+	media_batch_size = inline_data_size + 6;
+
+	cs = batch_alloc_items(batch, 8, media_batch_size);
+
+	*cs++ = MEDIA_OBJECT | (media_batch_size - 2);
+
+	/* interface descriptor offset */
+	*cs++ = 0;
+
+	/* without indirect data */
+	*cs++ = 0;
+	*cs++ = 0;
+
+	/* scoreboard */
+	*cs++ = 0;
+	*cs++ = 0;
+
+	/* inline */
+	*cs++ = (y_offset << 16) | (x_offset);
+	*cs++ = 0;
+	*cs++ = GT3_INLINE_DATA_DELAYS;
+	for (i = 3; i < inline_data_size; i++)
+		*cs++ = 0;
+
+	batch_advance(batch, cs);
+}
+
+static void gen7_emit_pipeline_flush(struct batch_chunk *batch)
+{
+	u32 *cs = batch_alloc_items(batch, 0, 5);
+
+	*cs++ = GFX_OP_PIPE_CONTROL(5);
+	*cs++ = PIPE_CONTROL_STATE_CACHE_INVALIDATE |
+		PIPE_CONTROL_GLOBAL_GTT_IVB;
+	*cs++ = 0;
+	*cs++ = 0;
+	*cs++ = 0;
+	batch_advance(batch, cs);
+}
+
+static void emit_batch(struct i915_vma * const vma,
+		       u32 *start,
+		       const struct batch_vals *bv)
+{
+	struct drm_i915_private *i915 = vma->vm->i915;
+	unsigned int desc_count = 64;
+	const u32 urb_size = 112;
+	struct batch_chunk cmds, state;
+	u32 interface_descriptor;
+	unsigned int i;
+
+	batch_init(&cmds, vma, start, 0, bv->cmd_size);
+	batch_init(&state, vma, start, bv->state_start, bv->state_size);
+
+	interface_descriptor =
+		gen7_fill_interface_descriptor(&state, bv,
+					       IS_HASWELL(i915) ?
+					       &cb_kernel_hsw :
+					       &cb_kernel_ivb,
+					       desc_count);
+	gen7_emit_pipeline_flush(&cmds);
+	batch_add(&cmds, PIPELINE_SELECT | PIPELINE_SELECT_MEDIA);
+	batch_add(&cmds, MI_NOOP);
+	gen7_emit_state_base_address(&cmds, interface_descriptor);
+	gen7_emit_pipeline_flush(&cmds);
+
+	gen7_emit_vfe_state(&cmds, bv, urb_size - 1, 0, 0);
+
+	gen7_emit_interface_descriptor_load(&cmds,
+					    interface_descriptor,
+					    desc_count);
+
+	for (i = 0; i < bv->max_primitives; i++)
+		gen7_emit_media_object(&cmds, i);
+
+	batch_add(&cmds, MI_BATCH_BUFFER_END);
+}
+
+int gen7_setup_clear_gpr_bb(struct intel_engine_cs * const engine,
+			    struct i915_vma * const vma)
+{
+	struct batch_vals bv;
+	u32 *batch;
+
+	batch_get_defaults(engine->i915, &bv);
+	if (!vma)
+		return bv.max_size;
+
+	GEM_BUG_ON(vma->obj->base.size < bv.max_size);
+
+	batch = i915_gem_object_pin_map(vma->obj, I915_MAP_WC);
+	if (IS_ERR(batch))
+		return PTR_ERR(batch);
+
+	emit_batch(vma, memset(batch, 0, bv.max_size), &bv);
+
+	i915_gem_object_flush_map(vma->obj);
+	i915_gem_object_unpin_map(vma->obj);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/gt/gen7_renderclear.h b/drivers/gpu/drm/i915/gt/gen7_renderclear.h
new file mode 100644
index 000000000000..bb100748e2c6
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/gen7_renderclear.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __GEN7_RENDERCLEAR_H__
+#define __GEN7_RENDERCLEAR_H__
+
+struct intel_engine_cs;
+struct i915_vma;
+
+int gen7_setup_clear_gpr_bb(struct intel_engine_cs * const engine,
+			    struct i915_vma * const vma);
+
+#endif /* __GEN7_RENDERCLEAR_H__ */
diff --git a/drivers/gpu/drm/i915/gt/gen8_ppgtt.c b/drivers/gpu/drm/i915/gt/gen8_ppgtt.c
index 4d1de2d97d5c..94e746af8926 100644
--- a/drivers/gpu/drm/i915/gt/gen8_ppgtt.c
+++ b/drivers/gpu/drm/i915/gt/gen8_ppgtt.c
@@ -8,6 +8,7 @@
 #include "gen8_ppgtt.h"
 #include "i915_scatterlist.h"
 #include "i915_trace.h"
+#include "i915_pvinfo.h"
 #include "i915_vgpu.h"
 #include "intel_gt.h"
 #include "intel_gtt.h"
@@ -25,6 +26,30 @@ static u64 gen8_pde_encode(const dma_addr_t addr,
 	return pde;
 }
 
+static u64 gen8_pte_encode(dma_addr_t addr,
+			   enum i915_cache_level level,
+			   u32 flags)
+{
+	gen8_pte_t pte = addr | _PAGE_PRESENT | _PAGE_RW;
+
+	if (unlikely(flags & PTE_READ_ONLY))
+		pte &= ~_PAGE_RW;
+
+	switch (level) {
+	case I915_CACHE_NONE:
+		pte |= PPAT_UNCACHED;
+		break;
+	case I915_CACHE_WT:
+		pte |= PPAT_DISPLAY_ELLC;
+		break;
+	default:
+		pte |= PPAT_CACHED;
+		break;
+	}
+
+	return pte;
+}
+
 static void gen8_ppgtt_notify_vgt(struct i915_ppgtt *ppgtt, bool create)
 {
 	struct drm_i915_private *i915 = ppgtt->vm.i915;
@@ -706,6 +731,8 @@ struct i915_ppgtt *gen8_ppgtt_create(struct intel_gt *gt)
 	ppgtt->vm.allocate_va_range = gen8_ppgtt_alloc;
 	ppgtt->vm.clear_range = gen8_ppgtt_clear;
 
+	ppgtt->vm.pte_encode = gen8_pte_encode;
+
 	if (intel_vgpu_active(gt->i915))
 		gen8_ppgtt_notify_vgt(ppgtt, true);
 
diff --git a/drivers/gpu/drm/i915/gt/hsw_clear_kernel.c b/drivers/gpu/drm/i915/gt/hsw_clear_kernel.c
new file mode 100644
index 000000000000..b47f9d4a0848
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/hsw_clear_kernel.c
@@ -0,0 +1,61 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2020 Intel Corporation
+ *
+ * Generated by: IGT Gpu Tools on Fri 21 Feb 2020 05:30:13 AM UTC
+ */
+
+static const u32 hsw_clear_kernel[] = {
+	0x00000001, 0x26020128, 0x00000024, 0x00000000,
+	0x00000040, 0x20280c21, 0x00000028, 0x00000001,
+	0x01000010, 0x20000c20, 0x0000002c, 0x00000000,
+	0x00010220, 0x34001c00, 0x00001400, 0x00000160,
+	0x00600001, 0x20600061, 0x00000000, 0x00000000,
+	0x00000008, 0x20601c85, 0x00000e00, 0x0000000c,
+	0x00000005, 0x20601ca5, 0x00000060, 0x00000001,
+	0x00000008, 0x20641c85, 0x00000e00, 0x0000000d,
+	0x00000005, 0x20641ca5, 0x00000064, 0x00000003,
+	0x00000041, 0x207424a5, 0x00000064, 0x00000034,
+	0x00000040, 0x206014a5, 0x00000060, 0x00000074,
+	0x00000008, 0x20681c85, 0x00000e00, 0x00000008,
+	0x00000005, 0x20681ca5, 0x00000068, 0x0000000f,
+	0x00000041, 0x20701ca5, 0x00000060, 0x00000010,
+	0x00000040, 0x206814a5, 0x00000068, 0x00000070,
+	0x00600001, 0x20a00061, 0x00000000, 0x00000000,
+	0x00000005, 0x206c1c85, 0x00000e00, 0x00000007,
+	0x00000041, 0x206c1ca5, 0x0000006c, 0x00000004,
+	0x00600001, 0x20800021, 0x008d0000, 0x00000000,
+	0x00000001, 0x20800021, 0x0000006c, 0x00000000,
+	0x00000001, 0x20840021, 0x00000068, 0x00000000,
+	0x00000001, 0x20880061, 0x00000000, 0x00000003,
+	0x00000005, 0x208c0d21, 0x00000086, 0xffffffff,
+	0x05600032, 0x20a00fa1, 0x008d0080, 0x02190001,
+	0x00000040, 0x20a01ca5, 0x000000a0, 0x00000001,
+	0x05600032, 0x20a00fa1, 0x008d0080, 0x040a8001,
+	0x02000040, 0x20281c21, 0x00000028, 0xffffffff,
+	0x00010220, 0x34001c00, 0x00001400, 0xffffffe0,
+	0x00000001, 0x26020128, 0x00000024, 0x00000000,
+	0x00000001, 0x220010e4, 0x00000000, 0x00000000,
+	0x00000001, 0x220831ec, 0x00000000, 0x007f007f,
+	0x00600001, 0x20400021, 0x008d0000, 0x00000000,
+	0x00600001, 0x2fe00021, 0x008d0000, 0x00000000,
+	0x00200001, 0x20400121, 0x00450020, 0x00000000,
+	0x00000001, 0x20480061, 0x00000000, 0x000f000f,
+	0x00000005, 0x204c0d21, 0x00000046, 0xffffffef,
+	0x00800001, 0x20600061, 0x00000000, 0x00000000,
+	0x00800001, 0x20800061, 0x00000000, 0x00000000,
+	0x00800001, 0x20a00061, 0x00000000, 0x00000000,
+	0x00800001, 0x20c00061, 0x00000000, 0x00000000,
+	0x00800001, 0x20e00061, 0x00000000, 0x00000000,
+	0x00800001, 0x21000061, 0x00000000, 0x00000000,
+	0x00800001, 0x21200061, 0x00000000, 0x00000000,
+	0x00800001, 0x21400061, 0x00000000, 0x00000000,
+	0x05600032, 0x20000fa0, 0x008d0040, 0x120a8000,
+	0x00000040, 0x20402d21, 0x00000020, 0x00100010,
+	0x05600032, 0x20000fa0, 0x008d0040, 0x120a8000,
+	0x02000040, 0x22083d8c, 0x00000208, 0xffffffff,
+	0x00800001, 0xa0000109, 0x00000602, 0x00000000,
+	0x00000040, 0x22001c84, 0x00000200, 0x00000020,
+	0x00010220, 0x34001c00, 0x00001400, 0xffffffc0,
+	0x07600032, 0x20000fa0, 0x008d0fe0, 0x82000010,
+};
diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c
index 8bb444cda14f..01474d3a558b 100644
--- a/drivers/gpu/drm/i915/gt/intel_context.c
+++ b/drivers/gpu/drm/i915/gt/intel_context.c
@@ -51,6 +51,11 @@ int intel_context_alloc_state(struct intel_context *ce)
 		return -EINTR;
 
 	if (!test_bit(CONTEXT_ALLOC_BIT, &ce->flags)) {
+		if (intel_context_is_banned(ce)) {
+			err = -EIO;
+			goto unlock;
+		}
+
 		err = ce->ops->alloc(ce);
 		if (unlikely(err))
 			goto unlock;
diff --git a/drivers/gpu/drm/i915/gt/intel_context_param.c b/drivers/gpu/drm/i915/gt/intel_context_param.c
new file mode 100644
index 000000000000..65dcd090245d
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_context_param.c
@@ -0,0 +1,63 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "i915_active.h"
+#include "intel_context.h"
+#include "intel_context_param.h"
+#include "intel_ring.h"
+
+int intel_context_set_ring_size(struct intel_context *ce, long sz)
+{
+	int err;
+
+	if (intel_context_lock_pinned(ce))
+		return -EINTR;
+
+	err = i915_active_wait(&ce->active);
+	if (err < 0)
+		goto unlock;
+
+	if (intel_context_is_pinned(ce)) {
+		err = -EBUSY; /* In active use, come back later! */
+		goto unlock;
+	}
+
+	if (test_bit(CONTEXT_ALLOC_BIT, &ce->flags)) {
+		struct intel_ring *ring;
+
+		/* Replace the existing ringbuffer */
+		ring = intel_engine_create_ring(ce->engine, sz);
+		if (IS_ERR(ring)) {
+			err = PTR_ERR(ring);
+			goto unlock;
+		}
+
+		intel_ring_put(ce->ring);
+		ce->ring = ring;
+
+		/* Context image will be updated on next pin */
+	} else {
+		ce->ring = __intel_context_ring_size(sz);
+	}
+
+unlock:
+	intel_context_unlock_pinned(ce);
+	return err;
+}
+
+long intel_context_get_ring_size(struct intel_context *ce)
+{
+	long sz = (unsigned long)READ_ONCE(ce->ring);
+
+	if (test_bit(CONTEXT_ALLOC_BIT, &ce->flags)) {
+		if (intel_context_lock_pinned(ce))
+			return -EINTR;
+
+		sz = ce->ring->size;
+		intel_context_unlock_pinned(ce);
+	}
+
+	return sz;
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_context_param.h b/drivers/gpu/drm/i915/gt/intel_context_param.h
new file mode 100644
index 000000000000..f053d8633fe2
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/intel_context_param.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_CONTEXT_PARAM_H
+#define INTEL_CONTEXT_PARAM_H
+
+struct intel_context;
+
+int intel_context_set_ring_size(struct intel_context *ce, long sz);
+long intel_context_get_ring_size(struct intel_context *ce);
+
+#endif /* INTEL_CONTEXT_PARAM_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h
index 11278343b9b5..0f3b68b95c56 100644
--- a/drivers/gpu/drm/i915/gt/intel_context_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_context_types.h
@@ -45,8 +45,8 @@ struct intel_context {
 
 	struct intel_engine_cs *engine;
 	struct intel_engine_cs *inflight;
-#define intel_context_inflight(ce) ptr_mask_bits((ce)->inflight, 2)
-#define intel_context_inflight_count(ce) ptr_unmask_bits((ce)->inflight, 2)
+#define intel_context_inflight(ce) ptr_mask_bits(READ_ONCE((ce)->inflight), 2)
+#define intel_context_inflight_count(ce) ptr_unmask_bits(READ_ONCE((ce)->inflight), 2)
 
 	struct i915_address_space *vm;
 	struct i915_gem_context __rcu *gem_context;
diff --git a/drivers/gpu/drm/i915/gt/intel_engine.h b/drivers/gpu/drm/i915/gt/intel_engine.h
index 29c8c03c5caa..b469de0dd9b6 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine.h
@@ -107,7 +107,20 @@ execlists_num_ports(const struct intel_engine_execlists * const execlists)
 static inline struct i915_request *
 execlists_active(const struct intel_engine_execlists *execlists)
 {
-	return *READ_ONCE(execlists->active);
+	struct i915_request * const *cur, * const *old, *active;
+
+	cur = READ_ONCE(execlists->active);
+	smp_rmb(); /* pairs with overwrite protection in process_csb() */
+	do {
+		old = cur;
+
+		active = READ_ONCE(*cur);
+		cur = READ_ONCE(execlists->active);
+
+		smp_rmb(); /* and complete the seqlock retry */
+	} while (unlikely(cur != old));
+
+	return active;
 }
 
 static inline void
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
index 119c9cb24fd4..3aa8a652c16d 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
@@ -275,6 +275,7 @@ static void intel_engine_sanitize_mmio(struct intel_engine_cs *engine)
 static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 {
 	const struct engine_info *info = &intel_engines[id];
+	struct drm_i915_private *i915 = gt->i915;
 	struct intel_engine_cs *engine;
 
 	BUILD_BUG_ON(MAX_ENGINE_CLASS >= BIT(GEN11_ENGINE_CLASS_WIDTH));
@@ -301,11 +302,11 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	engine->id = id;
 	engine->legacy_idx = INVALID_ENGINE;
 	engine->mask = BIT(id);
-	engine->i915 = gt->i915;
+	engine->i915 = i915;
 	engine->gt = gt;
 	engine->uncore = gt->uncore;
 	engine->hw_id = engine->guc_id = info->hw_id;
-	engine->mmio_base = __engine_mmio_base(gt->i915, info->mmio_bases);
+	engine->mmio_base = __engine_mmio_base(i915, info->mmio_bases);
 
 	engine->class = info->class;
 	engine->instance = info->instance;
@@ -313,6 +314,8 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 
 	engine->props.heartbeat_interval_ms =
 		CONFIG_DRM_I915_HEARTBEAT_INTERVAL;
+	engine->props.max_busywait_duration_ns =
+		CONFIG_DRM_I915_MAX_REQUEST_BUSYWAIT;
 	engine->props.preempt_timeout_ms =
 		CONFIG_DRM_I915_PREEMPT_TIMEOUT;
 	engine->props.stop_timeout_ms =
@@ -320,11 +323,15 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	engine->props.timeslice_duration_ms =
 		CONFIG_DRM_I915_TIMESLICE_DURATION;
 
+	/* Override to uninterruptible for OpenCL workloads. */
+	if (INTEL_GEN(i915) == 12 && engine->class == RENDER_CLASS)
+		engine->props.preempt_timeout_ms = 0;
+
 	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)
-		DRIVER_CAPS(gt->i915)->has_logical_contexts = true;
+		DRIVER_CAPS(i915)->has_logical_contexts = true;
 
 	/* Nothing to do here, execute in order of dependencies */
 	engine->schedule = NULL;
@@ -340,7 +347,7 @@ 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;
 
-	gt->i915->engine[id] = engine;
+	i915->engine[id] = engine;
 
 	return 0;
 }
@@ -1379,24 +1386,24 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine,
 			char hdr[160];
 			int len;
 
-			len = snprintf(hdr, sizeof(hdr),
-				       "\t\tActive[%d]: ",
-				       (int)(port - execlists->active));
+			len = scnprintf(hdr, sizeof(hdr),
+					"\t\tActive[%d]: ",
+					(int)(port - execlists->active));
 			if (!i915_request_signaled(rq)) {
 				struct intel_timeline *tl = get_timeline(rq);
 
-				len += snprintf(hdr + len, sizeof(hdr) - len,
-						"ring:{start:%08x, hwsp:%08x, seqno:%08x, runtime:%llums}, ",
-						i915_ggtt_offset(rq->ring->vma),
-						tl ? tl->hwsp_offset : 0,
-						hwsp_seqno(rq),
-						DIV_ROUND_CLOSEST_ULL(intel_context_get_total_runtime_ns(rq->context),
-								      1000 * 1000));
+				len += scnprintf(hdr + len, sizeof(hdr) - len,
+						 "ring:{start:%08x, hwsp:%08x, seqno:%08x, runtime:%llums}, ",
+						 i915_ggtt_offset(rq->ring->vma),
+						 tl ? tl->hwsp_offset : 0,
+						 hwsp_seqno(rq),
+						 DIV_ROUND_CLOSEST_ULL(intel_context_get_total_runtime_ns(rq->context),
+								       1000 * 1000));
 
 				if (tl)
 					intel_timeline_put(tl);
 			}
-			snprintf(hdr + len, sizeof(hdr) - len, "rq: ");
+			scnprintf(hdr + len, sizeof(hdr) - len, "rq: ");
 			print_request(m, rq, hdr);
 		}
 		for (port = execlists->pending; (rq = *port); port++) {
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h
index b23366a81048..80cdde712842 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h
@@ -547,6 +547,7 @@ struct intel_engine_cs {
 
 	struct {
 		unsigned long heartbeat_interval_ms;
+		unsigned long max_busywait_duration_ns;
 		unsigned long preempt_timeout_ms;
 		unsigned long stop_timeout_ms;
 		unsigned long timeslice_duration_ms;
diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt.c b/drivers/gpu/drm/i915/gt/intel_ggtt.c
index 41a00281f364..aed498a0d032 100644
--- a/drivers/gpu/drm/i915/gt/intel_ggtt.c
+++ b/drivers/gpu/drm/i915/gt/intel_ggtt.c
@@ -8,6 +8,8 @@
 #include <asm/set_memory.h>
 #include <asm/smp.h>
 
+#include <drm/i915_drm.h>
+
 #include "intel_gt.h"
 #include "i915_drv.h"
 #include "i915_scatterlist.h"
@@ -157,6 +159,13 @@ static void gmch_ggtt_invalidate(struct i915_ggtt *ggtt)
 	intel_gtt_chipset_flush();
 }
 
+static u64 gen8_ggtt_pte_encode(dma_addr_t addr,
+				enum i915_cache_level level,
+				u32 flags)
+{
+	return addr | _PAGE_PRESENT;
+}
+
 static void gen8_set_pte(void __iomem *addr, gen8_pte_t pte)
 {
 	writeq(pte, addr);
@@ -172,7 +181,7 @@ static void gen8_ggtt_insert_page(struct i915_address_space *vm,
 	gen8_pte_t __iomem *pte =
 		(gen8_pte_t __iomem *)ggtt->gsm + offset / I915_GTT_PAGE_SIZE;
 
-	gen8_set_pte(pte, gen8_pte_encode(addr, level, 0));
+	gen8_set_pte(pte, gen8_ggtt_pte_encode(addr, level, 0));
 
 	ggtt->invalidate(ggtt);
 }
@@ -185,7 +194,7 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm);
 	struct sgt_iter sgt_iter;
 	gen8_pte_t __iomem *gtt_entries;
-	const gen8_pte_t pte_encode = gen8_pte_encode(0, level, 0);
+	const gen8_pte_t pte_encode = gen8_ggtt_pte_encode(0, level, 0);
 	dma_addr_t addr;
 
 	/*
@@ -857,7 +866,7 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 	ggtt->vm.vma_ops.set_pages   = ggtt_set_pages;
 	ggtt->vm.vma_ops.clear_pages = clear_pages;
 
-	ggtt->vm.pte_encode = gen8_pte_encode;
+	ggtt->vm.pte_encode = gen8_ggtt_pte_encode;
 
 	setup_private_pat(ggtt->vm.gt->uncore);
 
diff --git a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
index 51b8718513bc..f04214a54f75 100644
--- a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
+++ b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
@@ -292,10 +292,21 @@
 #define MI_STORE_URB_MEM        MI_INSTR(0x2D, 0)
 #define MI_CONDITIONAL_BATCH_BUFFER_END MI_INSTR(0x36, 0)
 
-#define PIPELINE_SELECT                ((0x3<<29)|(0x1<<27)|(0x1<<24)|(0x4<<16))
-#define GFX_OP_3DSTATE_VF_STATISTICS   ((0x3<<29)|(0x1<<27)|(0x0<<24)|(0xB<<16))
-#define MEDIA_VFE_STATE                ((0x3<<29)|(0x2<<27)|(0x0<<24)|(0x0<<16))
+#define STATE_BASE_ADDRESS \
+	((0x3 << 29) | (0x0 << 27) | (0x1 << 24) | (0x1 << 16))
+#define BASE_ADDRESS_MODIFY		REG_BIT(0)
+#define PIPELINE_SELECT \
+	((0x3 << 29) | (0x1 << 27) | (0x1 << 24) | (0x4 << 16))
+#define PIPELINE_SELECT_MEDIA	       REG_BIT(0)
+#define GFX_OP_3DSTATE_VF_STATISTICS \
+	((0x3 << 29) | (0x1 << 27) | (0x0 << 24) | (0xB << 16))
+#define MEDIA_VFE_STATE \
+	((0x3 << 29) | (0x2 << 27) | (0x0 << 24) | (0x0 << 16))
 #define  MEDIA_VFE_STATE_MMIO_ACCESS_MASK (0x18)
+#define MEDIA_INTERFACE_DESCRIPTOR_LOAD \
+	((0x3 << 29) | (0x2 << 27) | (0x0 << 24) | (0x2 << 16))
+#define MEDIA_OBJECT \
+	((0x3 << 29) | (0x2 << 27) | (0x1 << 24) | (0x0 << 16))
 #define GPGPU_OBJECT                   ((0x3<<29)|(0x2<<27)|(0x1<<24)|(0x4<<16))
 #define GPGPU_WALKER                   ((0x3<<29)|(0x2<<27)|(0x1<<24)|(0x5<<16))
 #define GFX_OP_3DSTATE_DX9_CONSTANTF_VS \
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.c b/drivers/gpu/drm/i915/gt/intel_gt.c
index 3dea8881e915..d09f7596cb98 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt.c
@@ -667,6 +667,9 @@ void intel_gt_driver_release(struct intel_gt *gt)
 
 void intel_gt_driver_late_release(struct intel_gt *gt)
 {
+	/* We need to wait for inflight RCU frees to release their grip */
+	rcu_barrier();
+
 	intel_uc_driver_late_release(&gt->uc);
 	intel_gt_fini_requests(gt);
 	intel_gt_fini_reset(gt);
diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.c b/drivers/gpu/drm/i915/gt/intel_gtt.c
index bb9a6e638175..2a72cce63fd9 100644
--- a/drivers/gpu/drm/i915/gt/intel_gtt.c
+++ b/drivers/gpu/drm/i915/gt/intel_gtt.c
@@ -171,7 +171,9 @@ void __i915_vm_close(struct i915_address_space *vm)
 {
 	struct i915_vma *vma, *vn;
 
-	mutex_lock(&vm->mutex);
+	if (!atomic_dec_and_mutex_lock(&vm->open, &vm->mutex))
+		return;
+
 	list_for_each_entry_safe(vma, vn, &vm->bound_list, vm_link) {
 		struct drm_i915_gem_object *obj = vma->obj;
 
@@ -186,6 +188,7 @@ void __i915_vm_close(struct i915_address_space *vm)
 		i915_gem_object_put(obj);
 	}
 	GEM_BUG_ON(!list_empty(&vm->bound_list));
+
 	mutex_unlock(&vm->mutex);
 }
 
@@ -484,30 +487,6 @@ void gtt_write_workarounds(struct intel_gt *gt)
 	}
 }
 
-u64 gen8_pte_encode(dma_addr_t addr,
-		    enum i915_cache_level level,
-		    u32 flags)
-{
-	gen8_pte_t pte = addr | _PAGE_PRESENT | _PAGE_RW;
-
-	if (unlikely(flags & PTE_READ_ONLY))
-		pte &= ~_PAGE_RW;
-
-	switch (level) {
-	case I915_CACHE_NONE:
-		pte |= PPAT_UNCACHED;
-		break;
-	case I915_CACHE_WT:
-		pte |= PPAT_DISPLAY_ELLC;
-		break;
-	default:
-		pte |= PPAT_CACHED;
-		break;
-	}
-
-	return pte;
-}
-
 static void tgl_setup_private_ppat(struct intel_uncore *uncore)
 {
 	/* TGL doesn't support LLC or AGE settings */
diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.h b/drivers/gpu/drm/i915/gt/intel_gtt.h
index 23004445806a..b3116fe8d180 100644
--- a/drivers/gpu/drm/i915/gt/intel_gtt.h
+++ b/drivers/gpu/drm/i915/gt/intel_gtt.h
@@ -429,8 +429,7 @@ static inline void
 i915_vm_close(struct i915_address_space *vm)
 {
 	GEM_BUG_ON(!atomic_read(&vm->open));
-	if (atomic_dec_and_test(&vm->open))
-		__i915_vm_close(vm);
+	__i915_vm_close(vm);
 
 	i915_vm_put(vm);
 }
@@ -515,10 +514,6 @@ struct i915_ppgtt *i915_ppgtt_create(struct intel_gt *gt);
 void i915_ggtt_suspend(struct i915_ggtt *gtt);
 void i915_ggtt_resume(struct i915_ggtt *ggtt);
 
-u64 gen8_pte_encode(dma_addr_t addr,
-		    enum i915_cache_level level,
-		    u32 flags);
-
 int setup_page_dma(struct i915_address_space *vm, struct i915_page_dma *p);
 void cleanup_page_dma(struct i915_address_space *vm, struct i915_page_dma *p);
 
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index 47561dc29304..112531b29f59 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -245,7 +245,7 @@ static void mark_eio(struct i915_request *rq)
 
 	GEM_BUG_ON(i915_request_signaled(rq));
 
-	dma_fence_set_error(&rq->fence, -EIO);
+	i915_request_set_error_once(rq, -EIO);
 	i915_request_mark_complete(rq);
 }
 
@@ -293,7 +293,7 @@ static inline struct i915_priolist *to_priolist(struct rb_node *rb)
 
 static inline int rq_prio(const struct i915_request *rq)
 {
-	return rq->sched.attr.priority;
+	return READ_ONCE(rq->sched.attr.priority);
 }
 
 static int effective_prio(const struct i915_request *rq)
@@ -1004,7 +1004,7 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine)
 				i915_request_cancel_breadcrumb(rq);
 				spin_unlock(&rq->lock);
 			}
-			rq->engine = owner;
+			WRITE_ONCE(rq->engine, owner);
 			owner->submit_request(rq);
 			active = NULL;
 		}
@@ -1316,7 +1316,7 @@ __execlists_schedule_out(struct i915_request *rq,
 	 * If we have just completed this context, the engine may now be
 	 * idle and we want to re-enter powersaving.
 	 */
-	if (list_is_last(&rq->link, &ce->timeline->requests) &&
+	if (list_is_last_rcu(&rq->link, &ce->timeline->requests) &&
 	    i915_request_completed(rq))
 		intel_engine_add_retire(engine, ce->timeline);
 
@@ -1448,6 +1448,7 @@ assert_pending_valid(const struct intel_engine_execlists *execlists,
 {
 	struct i915_request * const *port, *rq;
 	struct intel_context *ce = NULL;
+	bool sentinel = false;
 
 	trace_ports(execlists, msg, execlists->pending);
 
@@ -1481,6 +1482,26 @@ assert_pending_valid(const struct intel_engine_execlists *execlists,
 		}
 		ce = rq->context;
 
+		/*
+		 * Sentinels are supposed to be lonely so they flush the
+		 * current exection off the HW. Check that they are the
+		 * only request in the pending submission.
+		 */
+		if (sentinel) {
+			GEM_TRACE_ERR("context:%llx after sentinel in pending[%zd]\n",
+				      ce->timeline->fence_context,
+				      port - execlists->pending);
+			return false;
+		}
+
+		sentinel = i915_request_has_sentinel(rq);
+		if (sentinel && port != execlists->pending) {
+			GEM_TRACE_ERR("sentinel context:%llx not in prime position[%zd]\n",
+				      ce->timeline->fence_context,
+				      port - execlists->pending);
+			return false;
+		}
+
 		/* Hold tightly onto the lock to prevent concurrent retires! */
 		if (!spin_trylock_irqsave(&rq->lock, flags))
 			continue;
@@ -1576,6 +1597,11 @@ static bool can_merge_ctx(const struct intel_context *prev,
 	return true;
 }
 
+static unsigned long i915_request_flags(const struct i915_request *rq)
+{
+	return READ_ONCE(rq->fence.flags);
+}
+
 static bool can_merge_rq(const struct i915_request *prev,
 			 const struct i915_request *next)
 {
@@ -1593,7 +1619,7 @@ static bool can_merge_rq(const struct i915_request *prev,
 	if (i915_request_completed(next))
 		return true;
 
-	if (unlikely((prev->fence.flags ^ next->fence.flags) &
+	if (unlikely((i915_request_flags(prev) ^ i915_request_flags(next)) &
 		     (BIT(I915_FENCE_FLAG_NOPREEMPT) |
 		      BIT(I915_FENCE_FLAG_SENTINEL))))
 		return false;
@@ -1601,6 +1627,7 @@ static bool can_merge_rq(const struct i915_request *prev,
 	if (!can_merge_ctx(prev->context, next->context))
 		return false;
 
+	GEM_BUG_ON(i915_seqno_passed(prev->fence.seqno, next->fence.seqno));
 	return true;
 }
 
@@ -1651,17 +1678,6 @@ static void virtual_xfer_breadcrumbs(struct virtual_engine *ve,
 	spin_unlock(&old->breadcrumbs.irq_lock);
 }
 
-static struct i915_request *
-last_active(const struct intel_engine_execlists *execlists)
-{
-	struct i915_request * const *last = READ_ONCE(execlists->active);
-
-	while (*last && i915_request_completed(*last))
-		last++;
-
-	return *last;
-}
-
 #define for_each_waiter(p__, rq__) \
 	list_for_each_entry_lockless(p__, \
 				     &(rq__)->sched.waiters_list, \
@@ -1735,11 +1751,9 @@ need_timeslice(struct intel_engine_cs *engine, const struct i915_request *rq)
 	if (!intel_engine_has_timeslices(engine))
 		return false;
 
-	if (list_is_last(&rq->sched.link, &engine->active.requests))
-		return false;
-
-	hint = max(rq_prio(list_next_entry(rq, sched.link)),
-		   engine->execlists.queue_priority_hint);
+	hint = engine->execlists.queue_priority_hint;
+	if (!list_is_last(&rq->sched.link, &engine->active.requests))
+		hint = max(hint, rq_prio(list_next_entry(rq, sched.link)));
 
 	return hint >= effective_prio(rq);
 }
@@ -1762,12 +1776,13 @@ timeslice(const struct intel_engine_cs *engine)
 static unsigned long
 active_timeslice(const struct intel_engine_cs *engine)
 {
-	const struct i915_request *rq = *engine->execlists.active;
+	const struct intel_engine_execlists *execlists = &engine->execlists;
+	const struct i915_request *rq = *execlists->active;
 
 	if (!rq || i915_request_completed(rq))
 		return 0;
 
-	if (engine->execlists.switch_priority_hint < effective_prio(rq))
+	if (READ_ONCE(execlists->switch_priority_hint) < effective_prio(rq))
 		return 0;
 
 	return timeslice(engine);
@@ -1781,16 +1796,29 @@ static void set_timeslice(struct intel_engine_cs *engine)
 	set_timer_ms(&engine->execlists.timer, active_timeslice(engine));
 }
 
+static void start_timeslice(struct intel_engine_cs *engine)
+{
+	struct intel_engine_execlists *execlists = &engine->execlists;
+	int prio = queue_prio(execlists);
+
+	WRITE_ONCE(execlists->switch_priority_hint, prio);
+	if (prio == INT_MIN)
+		return;
+
+	if (timer_pending(&execlists->timer))
+		return;
+
+	set_timer_ms(&execlists->timer, timeslice(engine));
+}
+
 static void record_preemption(struct intel_engine_execlists *execlists)
 {
 	(void)I915_SELFTEST_ONLY(execlists->preempt_hang.count++);
 }
 
-static unsigned long active_preempt_timeout(struct intel_engine_cs *engine)
+static unsigned long active_preempt_timeout(struct intel_engine_cs *engine,
+					    const struct i915_request *rq)
 {
-	struct i915_request *rq;
-
-	rq = last_active(&engine->execlists);
 	if (!rq)
 		return 0;
 
@@ -1801,13 +1829,14 @@ static unsigned long active_preempt_timeout(struct intel_engine_cs *engine)
 	return READ_ONCE(engine->props.preempt_timeout_ms);
 }
 
-static void set_preempt_timeout(struct intel_engine_cs *engine)
+static void set_preempt_timeout(struct intel_engine_cs *engine,
+				const struct i915_request *rq)
 {
 	if (!intel_engine_has_preempt_reset(engine))
 		return;
 
 	set_timer_ms(&engine->execlists.preempt,
-		     active_preempt_timeout(engine));
+		     active_preempt_timeout(engine, rq));
 }
 
 static inline void clear_ports(struct i915_request **ports, int count)
@@ -1820,6 +1849,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_request **port = execlists->pending;
 	struct i915_request ** const last_port = port + execlists->port_mask;
+	struct i915_request * const *active;
 	struct i915_request *last;
 	struct rb_node *rb;
 	bool submit = false;
@@ -1874,7 +1904,10 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	 * i.e. we will retrigger preemption following the ack in case
 	 * of trouble.
 	 */
-	last = last_active(execlists);
+	active = READ_ONCE(execlists->active);
+	while ((last = *active) && i915_request_completed(last))
+		active++;
+
 	if (last) {
 		if (need_preempt(engine, last, rb)) {
 			ENGINE_TRACE(engine,
@@ -1944,11 +1977,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				 * Even if ELSP[1] is occupied and not worthy
 				 * of timeslices, our queue might be.
 				 */
-				if (!execlists->timer.expires &&
-				    need_timeslice(engine, last))
-					set_timer_ms(&execlists->timer,
-						     timeslice(engine));
-
+				start_timeslice(engine);
 				return;
 			}
 		}
@@ -1983,7 +2012,8 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 
 			if (last && !can_merge_rq(last, rq)) {
 				spin_unlock(&ve->base.active.lock);
-				return; /* leave this for another */
+				start_timeslice(engine);
+				return; /* leave this for another sibling */
 			}
 
 			ENGINE_TRACE(engine,
@@ -1995,13 +2025,14 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				     "",
 				     yesno(engine != ve->siblings[0]));
 
-			ve->request = NULL;
-			ve->base.execlists.queue_priority_hint = INT_MIN;
+			WRITE_ONCE(ve->request, NULL);
+			WRITE_ONCE(ve->base.execlists.queue_priority_hint,
+				   INT_MIN);
 			rb_erase_cached(rb, &execlists->virtual);
 			RB_CLEAR_NODE(rb);
 
 			GEM_BUG_ON(!(rq->execution_mask & engine->mask));
-			rq->engine = engine;
+			WRITE_ONCE(rq->engine, engine);
 
 			if (engine != ve->siblings[0]) {
 				u32 *regs = ve->context.lrc_reg_state;
@@ -2121,6 +2152,9 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				GEM_BUG_ON(last &&
 					   !can_merge_ctx(last->context,
 							  rq->context));
+				GEM_BUG_ON(last &&
+					   i915_seqno_passed(last->fence.seqno,
+							     rq->fence.seqno));
 
 				submit = true;
 				last = rq;
@@ -2159,7 +2193,7 @@ done:
 		 * Skip if we ended up with exactly the same set of requests,
 		 * e.g. trying to timeslice a pair of ordered contexts
 		 */
-		if (!memcmp(execlists->active, execlists->pending,
+		if (!memcmp(active, execlists->pending,
 			    (port - execlists->pending + 1) * sizeof(*port))) {
 			do
 				execlists_schedule_out(fetch_and_zero(port));
@@ -2170,7 +2204,7 @@ done:
 		clear_ports(port + 1, last_port - port);
 
 		execlists_submit_ports(engine);
-		set_preempt_timeout(engine);
+		set_preempt_timeout(engine, *active);
 	} else {
 skip_submit:
 		ring_set_paused(engine, 0);
@@ -2191,6 +2225,7 @@ cancel_port_requests(struct intel_engine_execlists * const execlists)
 		execlists_schedule_out(*port);
 	clear_ports(execlists->inflight, ARRAY_SIZE(execlists->inflight));
 
+	smp_wmb(); /* complete the seqlock for execlists_active() */
 	WRITE_ONCE(execlists->active, execlists->inflight);
 }
 
@@ -2345,6 +2380,7 @@ static void process_csb(struct intel_engine_cs *engine)
 
 			/* Point active to the new ELSP; prevent overwriting */
 			WRITE_ONCE(execlists->active, execlists->pending);
+			smp_wmb(); /* notify execlists_active() */
 
 			/* cancel old inflight, prepare for switch */
 			trace_ports(execlists, "preempted", old);
@@ -2352,11 +2388,12 @@ static void process_csb(struct intel_engine_cs *engine)
 				execlists_schedule_out(*old++);
 
 			/* switch pending to inflight */
-			WRITE_ONCE(execlists->active,
-				   memcpy(execlists->inflight,
-					  execlists->pending,
-					  execlists_num_ports(execlists) *
-					  sizeof(*execlists->pending)));
+			memcpy(execlists->inflight,
+			       execlists->pending,
+			       execlists_num_ports(execlists) *
+			       sizeof(*execlists->pending));
+			smp_wmb(); /* complete the seqlock */
+			WRITE_ONCE(execlists->active, execlists->inflight);
 
 			WRITE_ONCE(execlists->pending[0], NULL);
 		} else {
@@ -2579,6 +2616,10 @@ static void __execlists_unhold(struct i915_request *rq)
 			struct i915_request *w =
 				container_of(p->waiter, typeof(*w), sched);
 
+			/* Propagate any change in error status */
+			if (rq->fence.error)
+				i915_request_set_error_once(w, rq->fence.error);
+
 			if (w->engine != rq->engine)
 				continue;
 
@@ -2966,6 +3007,7 @@ __execlists_update_reg_state(const struct intel_context *ce,
 	regs[CTX_RING_START] = i915_ggtt_offset(ring->vma);
 	regs[CTX_RING_HEAD] = head;
 	regs[CTX_RING_TAIL] = ring->tail;
+	regs[CTX_RING_CTL] = RING_CTL_SIZE(ring->size) | RING_VALID;
 
 	/* RPCS */
 	if (engine->class == RENDER_CLASS) {
@@ -3636,9 +3678,6 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 	if (!rq)
 		goto unwind;
 
-	/* We still have requests in-flight; the engine should be active */
-	GEM_BUG_ON(!intel_engine_pm_is_awake(engine));
-
 	ce = rq->context;
 	GEM_BUG_ON(!i915_vma_is_pinned(ce->state));
 
@@ -3648,8 +3687,12 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 		goto out_replay;
 	}
 
+	/* We still have requests in-flight; the engine should be active */
+	GEM_BUG_ON(!intel_engine_pm_is_awake(engine));
+
 	/* Context has requests still in-flight; it should not be idle! */
 	GEM_BUG_ON(i915_active_is_idle(&ce->active));
+
 	rq = active_request(ce->timeline, rq);
 	head = intel_ring_wrap(ce->ring, rq->head);
 	GEM_BUG_ON(head == ce->ring->tail);
@@ -3723,7 +3766,10 @@ static void execlists_reset_rewind(struct intel_engine_cs *engine, bool stalled)
 
 static void nop_submission_tasklet(unsigned long data)
 {
+	struct intel_engine_cs * const engine = (struct intel_engine_cs *)data;
+
 	/* The driver is wedged; don't process any more events. */
+	WRITE_ONCE(engine->execlists.queue_priority_hint, INT_MIN);
 }
 
 static void execlists_reset_cancel(struct intel_engine_cs *engine)
@@ -4119,26 +4165,6 @@ static int gen12_emit_flush_render(struct i915_request *request,
 
 		*cs++ = preparser_disable(false);
 		intel_ring_advance(request, cs);
-
-		/*
-		 * Wa_1604544889:tgl
-		 */
-		if (IS_TGL_REVID(request->i915, TGL_REVID_A0, TGL_REVID_A0)) {
-			flags = 0;
-			flags |= PIPE_CONTROL_CS_STALL;
-			flags |= PIPE_CONTROL_HDC_PIPELINE_FLUSH;
-
-			flags |= PIPE_CONTROL_STORE_DATA_INDEX;
-			flags |= PIPE_CONTROL_QW_WRITE;
-
-			cs = intel_ring_begin(request, 6);
-			if (IS_ERR(cs))
-				return PTR_ERR(cs);
-
-			cs = gen8_emit_pipe_control(cs, flags,
-						    LRC_PPHWSP_SCRATCH_ADDR);
-			intel_ring_advance(request, cs);
-		}
 	}
 
 	return 0;
@@ -4877,7 +4903,7 @@ static intel_engine_mask_t virtual_submission_mask(struct virtual_engine *ve)
 	mask = rq->execution_mask;
 	if (unlikely(!mask)) {
 		/* Invalid selection, submit to a random engine in error */
-		i915_request_skip(rq, -ENODEV);
+		i915_request_set_error_once(rq, -ENODEV);
 		mask = ve->siblings[0]->mask;
 	}
 
@@ -4891,7 +4917,7 @@ static intel_engine_mask_t virtual_submission_mask(struct virtual_engine *ve)
 static void virtual_submission_tasklet(unsigned long data)
 {
 	struct virtual_engine * const ve = (struct virtual_engine *)data;
-	const int prio = ve->base.execlists.queue_priority_hint;
+	const int prio = READ_ONCE(ve->base.execlists.queue_priority_hint);
 	intel_engine_mask_t mask;
 	unsigned int n;
 
@@ -5287,11 +5313,15 @@ void intel_execlists_show_requests(struct intel_engine_cs *engine,
 		show_request(m, last, "\t\tE ");
 	}
 
-	last = NULL;
-	count = 0;
+	if (execlists->switch_priority_hint != INT_MIN)
+		drm_printf(m, "\t\tSwitch priority hint: %d\n",
+			   READ_ONCE(execlists->switch_priority_hint));
 	if (execlists->queue_priority_hint != INT_MIN)
 		drm_printf(m, "\t\tQueue priority hint: %d\n",
-			   execlists->queue_priority_hint);
+			   READ_ONCE(execlists->queue_priority_hint));
+
+	last = NULL;
+	count = 0;
 	for (rb = rb_first_cached(&execlists->queue); rb; rb = rb_next(rb)) {
 		struct i915_priolist *p = rb_entry(rb, typeof(*p), node);
 		int i;
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6.c b/drivers/gpu/drm/i915/gt/intel_rc6.c
index bef132709854..66c07c32745c 100644
--- a/drivers/gpu/drm/i915/gt/intel_rc6.c
+++ b/drivers/gpu/drm/i915/gt/intel_rc6.c
@@ -7,6 +7,7 @@
 #include <linux/pm_runtime.h>
 
 #include "i915_drv.h"
+#include "i915_vgpu.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
 #include "intel_rc6.h"
@@ -319,10 +320,10 @@ static int vlv_rc6_init(struct intel_rc6 *rc6)
 		return PTR_ERR(pctx);
 	}
 
-	GEM_BUG_ON(range_overflows_t(u64,
-				     i915->dsm.start,
-				     pctx->stolen->start,
-				     U32_MAX));
+	GEM_BUG_ON(range_overflows_end_t(u64,
+					 i915->dsm.start,
+					 pctx->stolen->start,
+					 U32_MAX));
 	pctx_paddr = i915->dsm.start + pctx->stolen->start;
 	intel_uncore_write(uncore, VLV_PCBR, pctx_paddr);
 
diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c
index aef6ab58d7d9..8b170c1876b3 100644
--- a/drivers/gpu/drm/i915/gt/intel_reset.c
+++ b/drivers/gpu/drm/i915/gt/intel_reset.c
@@ -48,8 +48,10 @@ static void engine_skip_context(struct i915_request *rq)
 
 	lockdep_assert_held(&engine->active.lock);
 	list_for_each_entry_continue(rq, &engine->active.requests, sched.link)
-		if (rq->context == hung_ctx)
-			i915_request_skip(rq, -EIO);
+		if (rq->context == hung_ctx) {
+			i915_request_set_error_once(rq, -EIO);
+			__i915_request_skip(rq);
+		}
 }
 
 static void client_mark_guilty(struct i915_gem_context *ctx, bool banned)
@@ -92,13 +94,7 @@ static bool mark_guilty(struct i915_request *rq)
 		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;
-	}
+		return intel_context_is_banned(rq->context);
 
 	atomic_inc(&ctx->guilty_count);
 
@@ -154,11 +150,12 @@ void __i915_request_reset(struct i915_request *rq, bool guilty)
 
 	rcu_read_lock(); /* protect the GEM context */
 	if (guilty) {
-		i915_request_skip(rq, -EIO);
+		i915_request_set_error_once(rq, -EIO);
+		__i915_request_skip(rq);
 		if (mark_guilty(rq))
 			engine_skip_context(rq);
 	} else {
-		dma_fence_set_error(&rq->fence, -EAGAIN);
+		i915_request_set_error_once(rq, -EAGAIN);
 		mark_innocent(rq);
 	}
 	rcu_read_unlock();
@@ -785,7 +782,7 @@ static void nop_submit_request(struct i915_request *request)
 	unsigned long flags;
 
 	RQ_TRACE(request, "-EIO\n");
-	dma_fence_set_error(&request->fence, -EIO);
+	i915_request_set_error_once(request, -EIO);
 
 	spin_lock_irqsave(&engine->active.lock, flags);
 	__i915_request_submit(request);
diff --git a/drivers/gpu/drm/i915/gt/intel_ring_submission.c b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
index f70b903a98bc..1424582e4a9b 100644
--- a/drivers/gpu/drm/i915/gt/intel_ring_submission.c
+++ b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
@@ -29,11 +29,10 @@
 
 #include <linux/log2.h>
 
-#include <drm/i915_drm.h>
-
 #include "gem/i915_gem_context.h"
 
 #include "gen6_ppgtt.h"
+#include "gen7_renderclear.h"
 #include "i915_drv.h"
 #include "i915_trace.h"
 #include "intel_context.h"
@@ -897,9 +896,7 @@ static void reset_cancel(struct intel_engine_cs *engine)
 
 	/* Mark all submitted requests as skipped. */
 	list_for_each_entry(request, &engine->active.requests, sched.link) {
-		if (!i915_request_signaled(request))
-			dma_fence_set_error(&request->fence, -EIO);
-
+		i915_request_set_error_once(request, -EIO);
 		i915_request_mark_complete(request);
 	}
 
@@ -1360,7 +1357,9 @@ static int load_pd_dir(struct i915_request *rq,
 	return rq->engine->emit_flush(rq, EMIT_FLUSH);
 }
 
-static inline int mi_set_context(struct i915_request *rq, u32 flags)
+static inline int mi_set_context(struct i915_request *rq,
+				 struct intel_context *ce,
+				 u32 flags)
 {
 	struct drm_i915_private *i915 = rq->i915;
 	struct intel_engine_cs *engine = rq->engine;
@@ -1435,7 +1434,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 
 	*cs++ = MI_NOOP;
 	*cs++ = MI_SET_CONTEXT;
-	*cs++ = i915_ggtt_offset(rq->context->state) | flags;
+	*cs++ = i915_ggtt_offset(ce->state) | flags;
 	/*
 	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
 	 * WaMiSetContext_Hang:snb,ivb,vlv
@@ -1550,13 +1549,56 @@ static int switch_mm(struct i915_request *rq, struct i915_address_space *vm)
 	return rq->engine->emit_flush(rq, EMIT_INVALIDATE);
 }
 
+static int clear_residuals(struct i915_request *rq)
+{
+	struct intel_engine_cs *engine = rq->engine;
+	int ret;
+
+	ret = switch_mm(rq, vm_alias(engine->kernel_context->vm));
+	if (ret)
+		return ret;
+
+	if (engine->kernel_context->state) {
+		ret = mi_set_context(rq,
+				     engine->kernel_context,
+				     MI_MM_SPACE_GTT | MI_RESTORE_INHIBIT);
+		if (ret)
+			return ret;
+	}
+
+	ret = engine->emit_bb_start(rq,
+				    engine->wa_ctx.vma->node.start, 0,
+				    0);
+	if (ret)
+		return ret;
+
+	ret = engine->emit_flush(rq, EMIT_FLUSH);
+	if (ret)
+		return ret;
+
+	/* Always invalidate before the next switch_mm() */
+	return engine->emit_flush(rq, EMIT_INVALIDATE);
+}
+
 static int switch_context(struct i915_request *rq)
 {
+	struct intel_engine_cs *engine = rq->engine;
 	struct intel_context *ce = rq->context;
+	void **residuals = NULL;
 	int ret;
 
 	GEM_BUG_ON(HAS_EXECLISTS(rq->i915));
 
+	if (engine->wa_ctx.vma && ce != engine->kernel_context) {
+		if (engine->wa_ctx.vma->private != ce) {
+			ret = clear_residuals(rq);
+			if (ret)
+				return ret;
+
+			residuals = &engine->wa_ctx.vma->private;
+		}
+	}
+
 	ret = switch_mm(rq, vm_alias(ce->vm));
 	if (ret)
 		return ret;
@@ -1564,7 +1606,7 @@ static int switch_context(struct i915_request *rq)
 	if (ce->state) {
 		u32 flags;
 
-		GEM_BUG_ON(rq->engine->id != RCS0);
+		GEM_BUG_ON(engine->id != RCS0);
 
 		/* For resource streamer on HSW+ and power context elsewhere */
 		BUILD_BUG_ON(HSW_MI_RS_SAVE_STATE_EN != MI_SAVE_EXT_STATE_EN);
@@ -1576,7 +1618,7 @@ static int switch_context(struct i915_request *rq)
 		else
 			flags |= MI_RESTORE_INHIBIT;
 
-		ret = mi_set_context(rq, flags);
+		ret = mi_set_context(rq, ce, flags);
 		if (ret)
 			return ret;
 	}
@@ -1585,6 +1627,20 @@ static int switch_context(struct i915_request *rq)
 	if (ret)
 		return ret;
 
+	/*
+	 * Now past the point of no return, this request _will_ be emitted.
+	 *
+	 * Or at least this preamble will be emitted, the request may be
+	 * interrupted prior to submitting the user payload. If so, we
+	 * still submit the "empty" request in order to preserve global
+	 * state tracking such as this, our tracking of the current
+	 * dirty context.
+	 */
+	if (residuals) {
+		intel_context_put(*residuals);
+		*residuals = intel_context_get(ce);
+	}
+
 	return 0;
 }
 
@@ -1769,6 +1825,11 @@ static void ring_release(struct intel_engine_cs *engine)
 
 	intel_engine_cleanup_common(engine);
 
+	if (engine->wa_ctx.vma) {
+		intel_context_put(engine->wa_ctx.vma->private);
+		i915_vma_unpin_and_release(&engine->wa_ctx.vma, 0);
+	}
+
 	intel_ring_unpin(engine->legacy.ring);
 	intel_ring_put(engine->legacy.ring);
 
@@ -1916,6 +1977,64 @@ static void setup_vecs(struct intel_engine_cs *engine)
 	engine->emit_fini_breadcrumb = gen7_xcs_emit_breadcrumb;
 }
 
+static int gen7_ctx_switch_bb_setup(struct intel_engine_cs * const engine,
+				    struct i915_vma * const vma)
+{
+	return gen7_setup_clear_gpr_bb(engine, vma);
+}
+
+static int gen7_ctx_switch_bb_init(struct intel_engine_cs *engine)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int size;
+	int err;
+
+	size = gen7_ctx_switch_bb_setup(engine, NULL /* probe size */);
+	if (size <= 0)
+		return size;
+
+	size = ALIGN(size, PAGE_SIZE);
+	obj = i915_gem_object_create_internal(engine->i915, size);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	vma = i915_vma_instance(obj, engine->gt->vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_obj;
+	}
+
+	vma->private = intel_context_create(engine); /* dummy residuals */
+	if (IS_ERR(vma->private)) {
+		err = PTR_ERR(vma->private);
+		goto err_obj;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_HIGH);
+	if (err)
+		goto err_private;
+
+	err = i915_vma_sync(vma);
+	if (err)
+		goto err_unpin;
+
+	err = gen7_ctx_switch_bb_setup(engine, vma);
+	if (err)
+		goto err_unpin;
+
+	engine->wa_ctx.vma = vma;
+	return 0;
+
+err_unpin:
+	i915_vma_unpin(vma);
+err_private:
+	intel_context_put(vma->private);
+err_obj:
+	i915_gem_object_put(obj);
+	return err;
+}
+
 int intel_ring_submission_setup(struct intel_engine_cs *engine)
 {
 	struct intel_timeline *timeline;
@@ -1969,11 +2088,19 @@ int intel_ring_submission_setup(struct intel_engine_cs *engine)
 
 	GEM_BUG_ON(timeline->hwsp_ggtt != engine->status_page.vma);
 
+	if (IS_GEN(engine->i915, 7) && engine->class == RENDER_CLASS) {
+		err = gen7_ctx_switch_bb_init(engine);
+		if (err)
+			goto err_ring_unpin;
+	}
+
 	/* Finally, take ownership and responsibility for cleanup! */
 	engine->release = ring_release;
 
 	return 0;
 
+err_ring_unpin:
+	intel_ring_unpin(ring);
 err_ring:
 	intel_ring_put(ring);
 err_timeline_unpin:
@@ -1984,3 +2111,7 @@ err:
 	intel_engine_cleanup_common(engine);
 	return err;
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftest_ring_submission.c"
+#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c
index 30ae29b30f11..87f9638d2cbf 100644
--- a/drivers/gpu/drm/i915/gt/intel_rps.c
+++ b/drivers/gpu/drm/i915/gt/intel_rps.c
@@ -4,6 +4,8 @@
  * Copyright © 2019 Intel Corporation
  */
 
+#include <drm/i915_drm.h>
+
 #include "i915_drv.h"
 #include "intel_gt.h"
 #include "intel_gt_irq.h"
@@ -55,7 +57,7 @@ static u32 rps_pm_mask(struct intel_rps *rps, u8 val)
 	if (val < rps->max_freq_softlimit)
 		mask |= GEN6_PM_RP_UP_EI_EXPIRED | GEN6_PM_RP_UP_THRESHOLD;
 
-	mask &= rps->pm_events;
+	mask &= READ_ONCE(rps->pm_events);
 
 	return rps_pm_sanitize_mask(rps, ~mask);
 }
@@ -68,17 +70,19 @@ static void rps_reset_ei(struct intel_rps *rps)
 static void rps_enable_interrupts(struct intel_rps *rps)
 {
 	struct intel_gt *gt = rps_to_gt(rps);
+	u32 events;
 
 	rps_reset_ei(rps);
 
 	if (IS_VALLEYVIEW(gt->i915))
 		/* WaGsvRC0ResidencyMethod:vlv */
-		rps->pm_events = GEN6_PM_RP_UP_EI_EXPIRED;
+		events = GEN6_PM_RP_UP_EI_EXPIRED;
 	else
-		rps->pm_events = (GEN6_PM_RP_UP_THRESHOLD |
-				  GEN6_PM_RP_DOWN_THRESHOLD |
-				  GEN6_PM_RP_DOWN_TIMEOUT);
+		events = (GEN6_PM_RP_UP_THRESHOLD |
+			  GEN6_PM_RP_DOWN_THRESHOLD |
+			  GEN6_PM_RP_DOWN_TIMEOUT);
 
+	WRITE_ONCE(rps->pm_events, events);
 	spin_lock_irq(&gt->irq_lock);
 	gen6_gt_pm_enable_irq(gt, rps->pm_events);
 	spin_unlock_irq(&gt->irq_lock);
@@ -115,8 +119,7 @@ static void rps_disable_interrupts(struct intel_rps *rps)
 {
 	struct intel_gt *gt = rps_to_gt(rps);
 
-	rps->pm_events = 0;
-
+	WRITE_ONCE(rps->pm_events, 0);
 	set(gt->uncore, GEN6_PMINTRMSK, rps_pm_sanitize_mask(rps, ~0u));
 
 	spin_lock_irq(&gt->irq_lock);
@@ -642,7 +645,7 @@ void intel_rps_mark_interactive(struct intel_rps *rps, bool interactive)
 {
 	mutex_lock(&rps->power.mutex);
 	if (interactive) {
-		if (!rps->power.interactive++ && rps->active)
+		if (!rps->power.interactive++ && READ_ONCE(rps->active))
 			rps_set_power(rps, HIGH_POWER);
 	} else {
 		GEM_BUG_ON(!rps->power.interactive);
@@ -719,11 +722,15 @@ void intel_rps_unpark(struct intel_rps *rps)
 	 * performance, jump directly to RPe as our starting frequency.
 	 */
 	mutex_lock(&rps->lock);
-	rps->active = true;
+
+	WRITE_ONCE(rps->active, true);
+
 	freq = max(rps->cur_freq, rps->efficient_freq),
 	freq = clamp(freq, rps->min_freq_softlimit, rps->max_freq_softlimit);
 	intel_rps_set(rps, freq);
+
 	rps->last_adj = 0;
+
 	mutex_unlock(&rps->lock);
 
 	if (INTEL_GEN(rps_to_i915(rps)) >= 6)
@@ -743,7 +750,7 @@ void intel_rps_park(struct intel_rps *rps)
 	if (INTEL_GEN(i915) >= 6)
 		rps_disable_interrupts(rps);
 
-	rps->active = false;
+	WRITE_ONCE(rps->active, false);
 	if (rps->last_freq <= rps->idle_freq)
 		return;
 
@@ -767,10 +774,10 @@ void intel_rps_park(struct intel_rps *rps)
 
 void intel_rps_boost(struct i915_request *rq)
 {
-	struct intel_rps *rps = &rq->engine->gt->rps;
+	struct intel_rps *rps = &READ_ONCE(rq->engine)->gt->rps;
 	unsigned long flags;
 
-	if (i915_request_signaled(rq) || !rps->active)
+	if (i915_request_signaled(rq) || !READ_ONCE(rps->active))
 		return;
 
 	/* Serializes with i915_request_retire() */
@@ -1453,12 +1460,12 @@ static void rps_work(struct work_struct *work)
 	u32 pm_iir = 0;
 
 	spin_lock_irq(&gt->irq_lock);
-	pm_iir = fetch_and_zero(&rps->pm_iir);
+	pm_iir = fetch_and_zero(&rps->pm_iir) & READ_ONCE(rps->pm_events);
 	client_boost = atomic_read(&rps->num_waiters);
 	spin_unlock_irq(&gt->irq_lock);
 
 	/* Make sure we didn't queue anything we're not going to process. */
-	if ((pm_iir & rps->pm_events) == 0 && !client_boost)
+	if (!pm_iir && !client_boost)
 		goto out;
 
 	mutex_lock(&rps->lock);
@@ -1554,11 +1561,15 @@ void gen11_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
 void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
 {
 	struct intel_gt *gt = rps_to_gt(rps);
+	u32 events;
 
-	if (pm_iir & rps->pm_events) {
+	events = pm_iir & READ_ONCE(rps->pm_events);
+	if (events) {
 		spin_lock(&gt->irq_lock);
-		gen6_gt_pm_mask_irq(gt, pm_iir & rps->pm_events);
-		rps->pm_iir |= pm_iir & rps->pm_events;
+
+		gen6_gt_pm_mask_irq(gt, events);
+		rps->pm_iir |= events;
+
 		schedule_work(&rps->work);
 		spin_unlock(&gt->irq_lock);
 	}
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline.c b/drivers/gpu/drm/i915/gt/intel_timeline.c
index 54e1e55f3c81..91debbc97c9a 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline.c
+++ b/drivers/gpu/drm/i915/gt/intel_timeline.c
@@ -192,11 +192,15 @@ static void cacheline_release(struct intel_timeline_cacheline *cl)
 
 static void cacheline_free(struct intel_timeline_cacheline *cl)
 {
+	if (!i915_active_acquire_if_busy(&cl->active)) {
+		__idle_cacheline_free(cl);
+		return;
+	}
+
 	GEM_BUG_ON(ptr_test_bit(cl->vaddr, CACHELINE_FREE));
 	cl->vaddr = ptr_set_bit(cl->vaddr, CACHELINE_FREE);
 
-	if (i915_active_is_idle(&cl->active))
-		__idle_cacheline_free(cl);
+	i915_active_release(&cl->active);
 }
 
 int intel_timeline_init(struct intel_timeline *timeline,
diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c
index 06cef3c18f26..5176ad1a3976 100644
--- a/drivers/gpu/drm/i915/gt/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c
@@ -575,12 +575,29 @@ static void icl_ctx_workarounds_init(struct intel_engine_cs *engine,
 	/* allow headerless messages for preemptible GPGPU context */
 	WA_SET_BIT_MASKED(GEN10_SAMPLER_MODE,
 			  GEN11_SAMPLER_ENABLE_HEADLESS_MSG);
+
+	/* Wa_1604278689:icl,ehl */
+	wa_write(wal, IVB_FBC_RT_BASE, 0xFFFFFFFF & ~ILK_FBC_RT_VALID);
+	wa_write_masked_or(wal, IVB_FBC_RT_BASE_UPPER,
+			   0, /* write-only register; skip validation */
+			   0xFFFFFFFF);
+
+	/* Wa_1406306137:icl,ehl */
+	wa_masked_en(wal, GEN9_ROW_CHICKEN4, GEN11_DIS_PICK_2ND_EU);
 }
 
 static void tgl_ctx_workarounds_init(struct intel_engine_cs *engine,
 				     struct i915_wa_list *wal)
 {
-	/* Wa_1409142259:tgl */
+	/*
+	 * Wa_1409142259:tgl
+	 * Wa_1409347922:tgl
+	 * Wa_1409252684:tgl
+	 * Wa_1409217633:tgl
+	 * Wa_1409207793:tgl
+	 * Wa_1409178076:tgl
+	 * Wa_1408979724:tgl
+	 */
 	WA_SET_BIT_MASKED(GEN11_COMMON_SLICE_CHICKEN3,
 			  GEN12_DISABLE_CPS_AWARE_COLOR_PIPE);
 
@@ -593,6 +610,11 @@ static void tgl_ctx_workarounds_init(struct intel_engine_cs *engine,
 	 */
 	wa_add(wal, FF_MODE2, FF_MODE2_TDS_TIMER_MASK,
 	       FF_MODE2_TDS_TIMER_128, 0);
+
+	/* WaDisableGPGPUMidThreadPreemption:tgl */
+	WA_SET_FIELD_MASKED(GEN8_CS_CHICKEN1,
+			    GEN9_PREEMPT_GPGPU_LEVEL_MASK,
+			    GEN9_PREEMPT_GPGPU_THREAD_GROUP_LEVEL);
 }
 
 static void
@@ -898,11 +920,6 @@ icl_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal)
 			    SLICE_UNIT_LEVEL_CLKGATE,
 			    MSCUNIT_CLKGATE_DIS);
 
-	/* Wa_1406680159:icl */
-	wa_write_or(wal,
-		    SUBSLICE_UNIT_LEVEL_CLKGATE,
-		    GWUNIT_CLKGATE_DIS);
-
 	/* Wa_1406838659:icl (pre-prod) */
 	if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0))
 		wa_write_or(wal,
@@ -931,7 +948,7 @@ tgl_gt_workarounds_init(struct drm_i915_private *i915, struct i915_wa_list *wal)
 			    SUBSLICE_UNIT_LEVEL_CLKGATE2,
 			    CPSSUNIT_CLKGATE_DIS);
 
-	/* Wa_1409180338:tgl */
+	/* Wa_1607087056:tgl also know as BUG:1409180338 */
 	if (IS_TGL_REVID(i915, TGL_REVID_A0, TGL_REVID_A0))
 		wa_write_or(wal,
 			    SLICE_UNIT_LEVEL_CLKGATE,
@@ -1246,6 +1263,7 @@ static void tgl_whitelist_build(struct intel_engine_cs *engine)
 	case RENDER_CLASS:
 		/*
 		 * WaAllowPMDepthAndInvocationCountAccessFromUMD:tgl
+		 * Wa_1408556865:tgl
 		 *
 		 * This covers 4 registers which are next to one another :
 		 *   - PS_INVOCATION_COUNT
@@ -1259,6 +1277,9 @@ static void tgl_whitelist_build(struct intel_engine_cs *engine)
 
 		/* Wa_1808121037:tgl */
 		whitelist_reg(w, GEN7_COMMON_SLICE_CHICKEN1);
+
+		/* Wa_1806527549:tgl */
+		whitelist_reg(w, HIZ_CHICKEN);
 		break;
 	default:
 		break;
@@ -1325,19 +1346,21 @@ rcs_engine_wa_init(struct intel_engine_cs *engine, struct i915_wa_list *wal)
 	struct drm_i915_private *i915 = engine->i915;
 
 	if (IS_TGL_REVID(i915, TGL_REVID_A0, TGL_REVID_A0)) {
-		/* Wa_1606700617:tgl */
-		wa_masked_en(wal,
-			     GEN9_CS_DEBUG_MODE1,
-			     FF_DOP_CLOCK_GATE_DISABLE);
-
-		/* Wa_1607138336:tgl */
+		/*
+		 * Wa_1607138336:tgl
+		 * Wa_1607063988:tgl
+		 */
 		wa_write_or(wal,
 			    GEN9_CTX_PREEMPT_REG,
 			    GEN12_DISABLE_POSH_BUSY_FF_DOP_CG);
 
-		/* Wa_1607030317:tgl */
-		/* Wa_1607186500:tgl */
-		/* Wa_1607297627:tgl */
+		/*
+		 * Wa_1607030317:tgl
+		 * Wa_1607186500:tgl
+		 * Wa_1607297627:tgl there is 3 entries for this WA on BSpec, 2
+		 * of then says it is fixed on B0 the other one says it is
+		 * permanent
+		 */
 		wa_masked_en(wal,
 			     GEN6_RC_SLEEP_PSMI_CONTROL,
 			     GEN12_WAIT_FOR_EVENT_POWER_DOWN_DISABLE |
@@ -1356,10 +1379,29 @@ rcs_engine_wa_init(struct intel_engine_cs *engine, struct i915_wa_list *wal)
 			    GEN7_FF_THREAD_MODE,
 			    GEN12_FF_TESSELATION_DOP_GATE_DISABLE);
 
+		/*
+		 * Wa_1409085225:tgl
+		 * Wa_14010229206:tgl
+		 */
+		wa_masked_en(wal, GEN9_ROW_CHICKEN4, GEN12_DISABLE_TDL_PUSH);
+
+		/* Wa_1408615072:tgl */
+		wa_write_or(wal, UNSLICE_UNIT_LEVEL_CLKGATE2,
+			    VSUNIT_CLKGATE_DIS_TGL);
+	}
+
+	if (IS_TIGERLAKE(i915)) {
 		/* Wa_1606931601:tgl */
+		wa_masked_en(wal, GEN7_ROW_CHICKEN2, GEN12_DISABLE_EARLY_READ);
+
+		/* Wa_1409804808:tgl */
+		wa_masked_en(wal, GEN7_ROW_CHICKEN2,
+			     GEN12_PUSH_CONST_DEREF_HOLD_DIS);
+
+		/* Wa_1606700617:tgl */
 		wa_masked_en(wal,
-			     GEN7_ROW_CHICKEN2,
-			     GEN12_DISABLE_EARLY_READ);
+			     GEN9_CS_DEBUG_MODE1,
+			     FF_DOP_CLOCK_GATE_DISABLE);
 	}
 
 	if (IS_GEN(i915, 11)) {
@@ -1425,10 +1467,38 @@ rcs_engine_wa_init(struct intel_engine_cs *engine, struct i915_wa_list *wal)
 				   GEN11_SCRATCH2,
 				   GEN11_COHERENT_PARTIAL_WRITE_MERGE_ENABLE,
 				   0);
+
+		/* WaEnable32PlaneMode:icl */
+		wa_masked_en(wal, GEN9_CSFE_CHICKEN1_RCS,
+			     GEN11_ENABLE_32_PLANE_MODE);
+
+		/*
+		 * Wa_1408615072:icl,ehl  (vsunit)
+		 * Wa_1407596294:icl,ehl  (hsunit)
+		 */
+		wa_write_or(wal, UNSLICE_UNIT_LEVEL_CLKGATE,
+			    VSUNIT_CLKGATE_DIS | HSUNIT_CLKGATE_DIS);
+
+		/* Wa_1407352427:icl,ehl */
+		wa_write_or(wal, UNSLICE_UNIT_LEVEL_CLKGATE2,
+			    PSDUNIT_CLKGATE_DIS);
+
+		/* Wa_1406680159:icl,ehl */
+		wa_write_or(wal,
+			    SUBSLICE_UNIT_LEVEL_CLKGATE,
+			    GWUNIT_CLKGATE_DIS);
+
+		/*
+		 * Wa_1408767742:icl[a2..forever],ehl[all]
+		 * Wa_1605460711:icl[a0..c0]
+		 */
+		wa_write_or(wal,
+			    GEN7_FF_THREAD_MODE,
+			    GEN12_FF_TESSELATION_DOP_GATE_DISABLE);
 	}
 
-	if (IS_GEN_RANGE(i915, 9, 11)) {
-		/* FtrPerCtxtPreemptionGranularityControl:skl,bxt,kbl,cfl,cnl,icl */
+	if (IS_GEN_RANGE(i915, 9, 12)) {
+		/* FtrPerCtxtPreemptionGranularityControl:skl,bxt,kbl,cfl,cnl,icl,tgl */
 		wa_masked_en(wal,
 			     GEN7_FF_SLICE_CS_CHICKEN1,
 			     GEN9_FFSC_PERCTX_PREEMPT_CTRL);
@@ -1595,15 +1665,34 @@ err_obj:
 	return ERR_PTR(err);
 }
 
+static const struct {
+	u32 start;
+	u32 end;
+} mcr_ranges_gen8[] = {
+	{ .start = 0x5500, .end = 0x55ff },
+	{ .start = 0x7000, .end = 0x7fff },
+	{ .start = 0x9400, .end = 0x97ff },
+	{ .start = 0xb000, .end = 0xb3ff },
+	{ .start = 0xe000, .end = 0xe7ff },
+	{},
+};
+
 static bool mcr_range(struct drm_i915_private *i915, u32 offset)
 {
+	int i;
+
+	if (INTEL_GEN(i915) < 8)
+		return false;
+
 	/*
-	 * Registers in this range are affected by the MCR selector
+	 * Registers in these ranges 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.
 	 */
-	if (INTEL_GEN(i915) >= 8 && (offset >= 0xb000 && offset <= 0xb4ff))
-		return true;
+	for (i = 0; mcr_ranges_gen8[i].start; i++)
+		if (offset >= mcr_ranges_gen8[i].start &&
+		    offset <= mcr_ranges_gen8[i].end)
+			return true;
 
 	return false;
 }
diff --git a/drivers/gpu/drm/i915/gt/ivb_clear_kernel.c b/drivers/gpu/drm/i915/gt/ivb_clear_kernel.c
new file mode 100644
index 000000000000..610ca7687735
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/ivb_clear_kernel.c
@@ -0,0 +1,61 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2020 Intel Corporation
+ *
+ * Generated by: IGT Gpu Tools on Fri 21 Feb 2020 05:29:32 AM UTC
+ */
+
+static const u32 ivb_clear_kernel[] = {
+	0x00000001, 0x26020128, 0x00000024, 0x00000000,
+	0x00000040, 0x20280c21, 0x00000028, 0x00000001,
+	0x01000010, 0x20000c20, 0x0000002c, 0x00000000,
+	0x00010220, 0x34001c00, 0x00001400, 0x0000002c,
+	0x00600001, 0x20600061, 0x00000000, 0x00000000,
+	0x00000008, 0x20601c85, 0x00000e00, 0x0000000c,
+	0x00000005, 0x20601ca5, 0x00000060, 0x00000001,
+	0x00000008, 0x20641c85, 0x00000e00, 0x0000000d,
+	0x00000005, 0x20641ca5, 0x00000064, 0x00000003,
+	0x00000041, 0x207424a5, 0x00000064, 0x00000034,
+	0x00000040, 0x206014a5, 0x00000060, 0x00000074,
+	0x00000008, 0x20681c85, 0x00000e00, 0x00000008,
+	0x00000005, 0x20681ca5, 0x00000068, 0x0000000f,
+	0x00000041, 0x20701ca5, 0x00000060, 0x00000010,
+	0x00000040, 0x206814a5, 0x00000068, 0x00000070,
+	0x00600001, 0x20a00061, 0x00000000, 0x00000000,
+	0x00000005, 0x206c1c85, 0x00000e00, 0x00000007,
+	0x00000041, 0x206c1ca5, 0x0000006c, 0x00000004,
+	0x00600001, 0x20800021, 0x008d0000, 0x00000000,
+	0x00000001, 0x20800021, 0x0000006c, 0x00000000,
+	0x00000001, 0x20840021, 0x00000068, 0x00000000,
+	0x00000001, 0x20880061, 0x00000000, 0x00000003,
+	0x00000005, 0x208c0d21, 0x00000086, 0xffffffff,
+	0x05600032, 0x20a00fa1, 0x008d0080, 0x02190001,
+	0x00000040, 0x20a01ca5, 0x000000a0, 0x00000001,
+	0x05600032, 0x20a00fa1, 0x008d0080, 0x040a8001,
+	0x02000040, 0x20281c21, 0x00000028, 0xffffffff,
+	0x00010220, 0x34001c00, 0x00001400, 0xfffffffc,
+	0x00000001, 0x26020128, 0x00000024, 0x00000000,
+	0x00000001, 0x220010e4, 0x00000000, 0x00000000,
+	0x00000001, 0x220831ec, 0x00000000, 0x007f007f,
+	0x00600001, 0x20400021, 0x008d0000, 0x00000000,
+	0x00600001, 0x2fe00021, 0x008d0000, 0x00000000,
+	0x00200001, 0x20400121, 0x00450020, 0x00000000,
+	0x00000001, 0x20480061, 0x00000000, 0x000f000f,
+	0x00000005, 0x204c0d21, 0x00000046, 0xffffffef,
+	0x00800001, 0x20600061, 0x00000000, 0x00000000,
+	0x00800001, 0x20800061, 0x00000000, 0x00000000,
+	0x00800001, 0x20a00061, 0x00000000, 0x00000000,
+	0x00800001, 0x20c00061, 0x00000000, 0x00000000,
+	0x00800001, 0x20e00061, 0x00000000, 0x00000000,
+	0x00800001, 0x21000061, 0x00000000, 0x00000000,
+	0x00800001, 0x21200061, 0x00000000, 0x00000000,
+	0x00800001, 0x21400061, 0x00000000, 0x00000000,
+	0x05600032, 0x20000fa0, 0x008d0040, 0x120a8000,
+	0x00000040, 0x20402d21, 0x00000020, 0x00100010,
+	0x05600032, 0x20000fa0, 0x008d0040, 0x120a8000,
+	0x02000040, 0x22083d8c, 0x00000208, 0xffffffff,
+	0x00800001, 0xa0000109, 0x00000602, 0x00000000,
+	0x00000040, 0x22001c84, 0x00000200, 0x00000020,
+	0x00010220, 0x34001c00, 0x00001400, 0xfffffff8,
+	0x07600032, 0x20000fa0, 0x008d0fe0, 0x82000010,
+};
diff --git a/drivers/gpu/drm/i915/gt/mock_engine.c b/drivers/gpu/drm/i915/gt/mock_engine.c
index 5633515c12e9..4a53ded7c2dd 100644
--- a/drivers/gpu/drm/i915/gt/mock_engine.c
+++ b/drivers/gpu/drm/i915/gt/mock_engine.c
@@ -244,9 +244,7 @@ static void mock_reset_cancel(struct intel_engine_cs *engine)
 
 	/* Mark all submitted requests as skipped. */
 	list_for_each_entry(request, &engine->active.requests, sched.link) {
-		if (!i915_request_signaled(request))
-			dma_fence_set_error(&request->fence, -EIO);
-
+		i915_request_set_error_once(request, -EIO);
 		i915_request_mark_complete(request);
 	}
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c b/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
index 43d4d589749f..697114dd1f47 100644
--- a/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
+++ b/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
@@ -142,6 +142,24 @@ out:
 	return err;
 }
 
+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_idle_flush(void *arg)
 {
 	struct intel_gt *gt = arg;
@@ -152,9 +170,11 @@ static int live_idle_flush(void *arg)
 	/* Check that we can flush the idle barriers */
 
 	for_each_engine(engine, gt, id) {
-		intel_engine_pm_get(engine);
+		unsigned long heartbeat;
+
+		engine_heartbeat_disable(engine, &heartbeat);
 		err = __live_idle_pulse(engine, intel_engine_flush_barriers);
-		intel_engine_pm_put(engine);
+		engine_heartbeat_enable(engine, heartbeat);
 		if (err)
 			break;
 	}
@@ -172,9 +192,11 @@ static int live_idle_pulse(void *arg)
 	/* Check that heartbeat pulses flush the idle barriers */
 
 	for_each_engine(engine, gt, id) {
-		intel_engine_pm_get(engine);
+		unsigned long heartbeat;
+
+		engine_heartbeat_disable(engine, &heartbeat);
 		err = __live_idle_pulse(engine, intel_engine_pulse);
-		intel_engine_pm_put(engine);
+		engine_heartbeat_enable(engine, heartbeat);
 		if (err && err != -ENODEV)
 			break;
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
index c3514ec7b8db..2b2efff6e19d 100644
--- a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
+++ b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
@@ -268,7 +268,7 @@ hang_create_request(struct hang *h, struct intel_engine_cs *engine)
 
 cancel_rq:
 	if (err) {
-		i915_request_skip(rq, err);
+		i915_request_set_error_once(rq, err);
 		i915_request_add(rq);
 	}
 unpin_hws:
diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c
index febd608c23a7..6f06ba750a0a 100644
--- a/drivers/gpu/drm/i915/gt/selftest_lrc.c
+++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c
@@ -90,6 +90,49 @@ static int wait_for_submit(struct intel_engine_cs *engine,
 	return -ETIME;
 }
 
+static int wait_for_reset(struct intel_engine_cs *engine,
+			  struct i915_request *rq,
+			  unsigned long timeout)
+{
+	timeout += jiffies;
+
+	do {
+		cond_resched();
+		intel_engine_flush_submission(engine);
+
+		if (READ_ONCE(engine->execlists.pending[0]))
+			continue;
+
+		if (i915_request_completed(rq))
+			break;
+
+		if (READ_ONCE(rq->fence.error))
+			break;
+	} while (time_before(jiffies, timeout));
+
+	flush_scheduled_work();
+
+	if (rq->fence.error != -EIO) {
+		pr_err("%s: hanging request %llx:%lld not reset\n",
+		       engine->name,
+		       rq->fence.context,
+		       rq->fence.seqno);
+		return -EINVAL;
+	}
+
+	/* Give the request a jiffie to complete after flushing the worker */
+	if (i915_request_wait(rq, 0,
+			      max(0l, (long)(timeout - jiffies)) + 1) < 0) {
+		pr_err("%s: hanging request %llx:%lld did not complete\n",
+		       engine->name,
+		       rq->fence.context,
+		       rq->fence.seqno);
+		return -ETIME;
+	}
+
+	return 0;
+}
+
 static int live_sanitycheck(void *arg)
 {
 	struct intel_gt *gt = arg;
@@ -1805,14 +1848,9 @@ static int __cancel_active0(struct live_preempt_cancel *arg)
 	if (err)
 		goto out;
 
-	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
-		err = -EIO;
-		goto out;
-	}
-
-	if (rq->fence.error != -EIO) {
-		pr_err("Cancelled inflight0 request did not report -EIO\n");
-		err = -EINVAL;
+	err = wait_for_reset(arg->engine, rq, HZ / 2);
+	if (err) {
+		pr_err("Cancelled inflight0 request did not reset\n");
 		goto out;
 	}
 
@@ -1870,10 +1908,9 @@ static int __cancel_active1(struct live_preempt_cancel *arg)
 		goto out;
 
 	igt_spinner_end(&arg->a.spin);
-	if (i915_request_wait(rq[1], 0, HZ / 5) < 0) {
-		err = -EIO;
+	err = wait_for_reset(arg->engine, rq[1], HZ / 2);
+	if (err)
 		goto out;
-	}
 
 	if (rq[0]->fence.error != 0) {
 		pr_err("Normal inflight0 request did not complete\n");
@@ -1953,10 +1990,9 @@ static int __cancel_queued(struct live_preempt_cancel *arg)
 	if (err)
 		goto out;
 
-	if (i915_request_wait(rq[2], 0, HZ / 5) < 0) {
-		err = -EIO;
+	err = wait_for_reset(arg->engine, rq[2], HZ / 2);
+	if (err)
 		goto out;
-	}
 
 	if (rq[0]->fence.error != -EIO) {
 		pr_err("Cancelled inflight0 request did not report -EIO\n");
@@ -2014,14 +2050,9 @@ static int __cancel_hostile(struct live_preempt_cancel *arg)
 	if (err)
 		goto out;
 
-	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
-		err = -EIO;
-		goto out;
-	}
-
-	if (rq->fence.error != -EIO) {
-		pr_err("Cancelled inflight0 request did not report -EIO\n");
-		err = -EINVAL;
+	err = wait_for_reset(arg->engine, rq, HZ / 2);
+	if (err) {
+		pr_err("Cancelled inflight0 request did not reset\n");
 		goto out;
 	}
 
@@ -4015,6 +4046,32 @@ static int emit_semaphore_signal(struct intel_context *ce, void *slot)
 	return 0;
 }
 
+static int context_flush(struct intel_context *ce, long timeout)
+{
+	struct i915_request *rq;
+	struct dma_fence *fence;
+	int err = 0;
+
+	rq = intel_engine_create_kernel_request(ce->engine);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	fence = i915_active_fence_get(&ce->timeline->last_request);
+	if (fence) {
+		i915_request_await_dma_fence(rq, fence);
+		dma_fence_put(fence);
+	}
+
+	rq = i915_request_get(rq);
+	i915_request_add(rq);
+	if (i915_request_wait(rq, 0, timeout) < 0)
+		err = -ETIME;
+	i915_request_put(rq);
+
+	rmb(); /* We know the request is written, make sure all state is too! */
+	return err;
+}
+
 static int live_lrc_layout(void *arg)
 {
 	struct intel_gt *gt = arg;
@@ -4638,18 +4695,10 @@ static int __lrc_timestamp(const struct lrc_timestamp *arg, bool preempt)
 		wmb();
 	}
 
-	if (i915_request_wait(rq, 0, HZ / 2) < 0) {
-		err = -ETIME;
-		goto err;
-	}
-
-	/* and wait for switch to kernel */
-	if (igt_flush_test(arg->engine->i915)) {
-		err = -EIO;
+	/* And wait for switch to kernel (to save our context to memory) */
+	err = context_flush(arg->ce[0], HZ / 2);
+	if (err)
 		goto err;
-	}
-
-	rmb();
 
 	if (!timestamp_advanced(arg->poison, slot[1])) {
 		pr_err("%s(%s): invalid timestamp on restore, context:%x, request:%x\n",
@@ -4674,9 +4723,9 @@ err:
 
 static int live_lrc_timestamp(void *arg)
 {
+	struct lrc_timestamp data = {};
 	struct intel_gt *gt = arg;
 	enum intel_engine_id id;
-	struct lrc_timestamp data;
 	const u32 poison[] = {
 		0,
 		S32_MAX,
@@ -4748,6 +4797,677 @@ err:
 	return 0;
 }
 
+static struct i915_vma *
+create_user_vma(struct i915_address_space *vm, unsigned long size)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int err;
+
+	obj = i915_gem_object_create_internal(vm->i915, size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	vma = i915_vma_instance(obj, vm, NULL);
+	if (IS_ERR(vma)) {
+		i915_gem_object_put(obj);
+		return vma;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER);
+	if (err) {
+		i915_gem_object_put(obj);
+		return ERR_PTR(err);
+	}
+
+	return vma;
+}
+
+static struct i915_vma *
+store_context(struct intel_context *ce, struct i915_vma *scratch)
+{
+	struct i915_vma *batch;
+	u32 dw, x, *cs, *hw;
+
+	batch = create_user_vma(ce->vm, SZ_64K);
+	if (IS_ERR(batch))
+		return batch;
+
+	cs = i915_gem_object_pin_map(batch->obj, I915_MAP_WC);
+	if (IS_ERR(cs)) {
+		i915_vma_put(batch);
+		return ERR_CAST(cs);
+	}
+
+	x = 0;
+	dw = 0;
+	hw = ce->engine->pinned_default_state;
+	hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw);
+	do {
+		u32 len = hw[dw] & 0x7f;
+
+		if (hw[dw] == 0) {
+			dw++;
+			continue;
+		}
+
+		if ((hw[dw] & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) {
+			dw += len + 2;
+			continue;
+		}
+
+		dw++;
+		len = (len + 1) / 2;
+		while (len--) {
+			*cs++ = MI_STORE_REGISTER_MEM_GEN8;
+			*cs++ = hw[dw];
+			*cs++ = lower_32_bits(scratch->node.start + x);
+			*cs++ = upper_32_bits(scratch->node.start + x);
+
+			dw += 2;
+			x += 4;
+		}
+	} while (dw < PAGE_SIZE / sizeof(u32) &&
+		 (hw[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END);
+
+	*cs++ = MI_BATCH_BUFFER_END;
+
+	i915_gem_object_flush_map(batch->obj);
+	i915_gem_object_unpin_map(batch->obj);
+
+	return batch;
+}
+
+static int move_to_active(struct i915_request *rq,
+			  struct i915_vma *vma,
+			  unsigned int flags)
+{
+	int err;
+
+	i915_vma_lock(vma);
+	err = i915_request_await_object(rq, vma->obj, flags);
+	if (!err)
+		err = i915_vma_move_to_active(vma, rq, flags);
+	i915_vma_unlock(vma);
+
+	return err;
+}
+
+static struct i915_request *
+record_registers(struct intel_context *ce,
+		 struct i915_vma *before,
+		 struct i915_vma *after,
+		 u32 *sema)
+{
+	struct i915_vma *b_before, *b_after;
+	struct i915_request *rq;
+	u32 *cs;
+	int err;
+
+	b_before = store_context(ce, before);
+	if (IS_ERR(b_before))
+		return ERR_CAST(b_before);
+
+	b_after = store_context(ce, after);
+	if (IS_ERR(b_after)) {
+		rq = ERR_CAST(b_after);
+		goto err_before;
+	}
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		goto err_after;
+
+	err = move_to_active(rq, before, EXEC_OBJECT_WRITE);
+	if (err)
+		goto err_rq;
+
+	err = move_to_active(rq, b_before, 0);
+	if (err)
+		goto err_rq;
+
+	err = move_to_active(rq, after, EXEC_OBJECT_WRITE);
+	if (err)
+		goto err_rq;
+
+	err = move_to_active(rq, b_after, 0);
+	if (err)
+		goto err_rq;
+
+	cs = intel_ring_begin(rq, 14);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_rq;
+	}
+
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;
+	*cs++ = MI_BATCH_BUFFER_START_GEN8 | BIT(8);
+	*cs++ = lower_32_bits(b_before->node.start);
+	*cs++ = upper_32_bits(b_before->node.start);
+
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
+	*cs++ = MI_SEMAPHORE_WAIT |
+		MI_SEMAPHORE_GLOBAL_GTT |
+		MI_SEMAPHORE_POLL |
+		MI_SEMAPHORE_SAD_NEQ_SDD;
+	*cs++ = 0;
+	*cs++ = i915_ggtt_offset(ce->engine->status_page.vma) +
+		offset_in_page(sema);
+	*cs++ = 0;
+	*cs++ = MI_NOOP;
+
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;
+	*cs++ = MI_BATCH_BUFFER_START_GEN8 | BIT(8);
+	*cs++ = lower_32_bits(b_after->node.start);
+	*cs++ = upper_32_bits(b_after->node.start);
+
+	intel_ring_advance(rq, cs);
+
+	WRITE_ONCE(*sema, 0);
+	i915_request_get(rq);
+	i915_request_add(rq);
+err_after:
+	i915_vma_put(b_after);
+err_before:
+	i915_vma_put(b_before);
+	return rq;
+
+err_rq:
+	i915_request_add(rq);
+	rq = ERR_PTR(err);
+	goto err_after;
+}
+
+static struct i915_vma *load_context(struct intel_context *ce, u32 poison)
+{
+	struct i915_vma *batch;
+	u32 dw, *cs, *hw;
+
+	batch = create_user_vma(ce->vm, SZ_64K);
+	if (IS_ERR(batch))
+		return batch;
+
+	cs = i915_gem_object_pin_map(batch->obj, I915_MAP_WC);
+	if (IS_ERR(cs)) {
+		i915_vma_put(batch);
+		return ERR_CAST(cs);
+	}
+
+	dw = 0;
+	hw = ce->engine->pinned_default_state;
+	hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw);
+	do {
+		u32 len = hw[dw] & 0x7f;
+
+		if (hw[dw] == 0) {
+			dw++;
+			continue;
+		}
+
+		if ((hw[dw] & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) {
+			dw += len + 2;
+			continue;
+		}
+
+		dw++;
+		len = (len + 1) / 2;
+		*cs++ = MI_LOAD_REGISTER_IMM(len);
+		while (len--) {
+			*cs++ = hw[dw];
+			*cs++ = poison;
+			dw += 2;
+		}
+	} while (dw < PAGE_SIZE / sizeof(u32) &&
+		 (hw[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END);
+
+	*cs++ = MI_BATCH_BUFFER_END;
+
+	i915_gem_object_flush_map(batch->obj);
+	i915_gem_object_unpin_map(batch->obj);
+
+	return batch;
+}
+
+static int poison_registers(struct intel_context *ce, u32 poison, u32 *sema)
+{
+	struct i915_request *rq;
+	struct i915_vma *batch;
+	u32 *cs;
+	int err;
+
+	batch = load_context(ce, poison);
+	if (IS_ERR(batch))
+		return PTR_ERR(batch);
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_batch;
+	}
+
+	err = move_to_active(rq, batch, 0);
+	if (err)
+		goto err_rq;
+
+	cs = intel_ring_begin(rq, 8);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_rq;
+	}
+
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;
+	*cs++ = MI_BATCH_BUFFER_START_GEN8 | BIT(8);
+	*cs++ = lower_32_bits(batch->node.start);
+	*cs++ = upper_32_bits(batch->node.start);
+
+	*cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
+	*cs++ = i915_ggtt_offset(ce->engine->status_page.vma) +
+		offset_in_page(sema);
+	*cs++ = 0;
+	*cs++ = 1;
+
+	intel_ring_advance(rq, cs);
+
+	rq->sched.attr.priority = I915_PRIORITY_BARRIER;
+err_rq:
+	i915_request_add(rq);
+err_batch:
+	i915_vma_put(batch);
+	return err;
+}
+
+static bool is_moving(u32 a, u32 b)
+{
+	return a != b;
+}
+
+static int compare_isolation(struct intel_engine_cs *engine,
+			     struct i915_vma *ref[2],
+			     struct i915_vma *result[2],
+			     struct intel_context *ce,
+			     u32 poison)
+{
+	u32 x, dw, *hw, *lrc;
+	u32 *A[2], *B[2];
+	int err = 0;
+
+	A[0] = i915_gem_object_pin_map(ref[0]->obj, I915_MAP_WC);
+	if (IS_ERR(A[0]))
+		return PTR_ERR(A[0]);
+
+	A[1] = i915_gem_object_pin_map(ref[1]->obj, I915_MAP_WC);
+	if (IS_ERR(A[1])) {
+		err = PTR_ERR(A[1]);
+		goto err_A0;
+	}
+
+	B[0] = i915_gem_object_pin_map(result[0]->obj, I915_MAP_WC);
+	if (IS_ERR(B[0])) {
+		err = PTR_ERR(B[0]);
+		goto err_A1;
+	}
+
+	B[1] = i915_gem_object_pin_map(result[1]->obj, I915_MAP_WC);
+	if (IS_ERR(B[1])) {
+		err = PTR_ERR(B[1]);
+		goto err_B0;
+	}
+
+	lrc = i915_gem_object_pin_map(ce->state->obj,
+				      i915_coherent_map_type(engine->i915));
+	if (IS_ERR(lrc)) {
+		err = PTR_ERR(lrc);
+		goto err_B1;
+	}
+	lrc += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw);
+
+	x = 0;
+	dw = 0;
+	hw = engine->pinned_default_state;
+	hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw);
+	do {
+		u32 len = hw[dw] & 0x7f;
+
+		if (hw[dw] == 0) {
+			dw++;
+			continue;
+		}
+
+		if ((hw[dw] & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) {
+			dw += len + 2;
+			continue;
+		}
+
+		dw++;
+		len = (len + 1) / 2;
+		while (len--) {
+			if (!is_moving(A[0][x], A[1][x]) &&
+			    (A[0][x] != B[0][x] || A[1][x] != B[1][x])) {
+				switch (hw[dw] & 4095) {
+				case 0x30: /* RING_HEAD */
+				case 0x34: /* RING_TAIL */
+					break;
+
+				default:
+					pr_err("%s[%d]: Mismatch for register %4x, default %08x, reference %08x, result (%08x, %08x), poison %08x, context %08x\n",
+					       engine->name, dw,
+					       hw[dw], hw[dw + 1],
+					       A[0][x], B[0][x], B[1][x],
+					       poison, lrc[dw + 1]);
+					err = -EINVAL;
+					break;
+				}
+			}
+			dw += 2;
+			x++;
+		}
+	} while (dw < PAGE_SIZE / sizeof(u32) &&
+		 (hw[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END);
+
+	i915_gem_object_unpin_map(ce->state->obj);
+err_B1:
+	i915_gem_object_unpin_map(result[1]->obj);
+err_B0:
+	i915_gem_object_unpin_map(result[0]->obj);
+err_A1:
+	i915_gem_object_unpin_map(ref[1]->obj);
+err_A0:
+	i915_gem_object_unpin_map(ref[0]->obj);
+	return err;
+}
+
+static int __lrc_isolation(struct intel_engine_cs *engine, u32 poison)
+{
+	u32 *sema = memset32(engine->status_page.addr + 1000, 0, 1);
+	struct i915_vma *ref[2], *result[2];
+	struct intel_context *A, *B;
+	struct i915_request *rq;
+	int err;
+
+	A = intel_context_create(engine);
+	if (IS_ERR(A))
+		return PTR_ERR(A);
+
+	B = intel_context_create(engine);
+	if (IS_ERR(B)) {
+		err = PTR_ERR(B);
+		goto err_A;
+	}
+
+	ref[0] = create_user_vma(A->vm, SZ_64K);
+	if (IS_ERR(ref[0])) {
+		err = PTR_ERR(ref[0]);
+		goto err_B;
+	}
+
+	ref[1] = create_user_vma(A->vm, SZ_64K);
+	if (IS_ERR(ref[1])) {
+		err = PTR_ERR(ref[1]);
+		goto err_ref0;
+	}
+
+	rq = record_registers(A, ref[0], ref[1], sema);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_ref1;
+	}
+
+	WRITE_ONCE(*sema, 1);
+	wmb();
+
+	if (i915_request_wait(rq, 0, HZ / 2) < 0) {
+		i915_request_put(rq);
+		err = -ETIME;
+		goto err_ref1;
+	}
+	i915_request_put(rq);
+
+	result[0] = create_user_vma(A->vm, SZ_64K);
+	if (IS_ERR(result[0])) {
+		err = PTR_ERR(result[0]);
+		goto err_ref1;
+	}
+
+	result[1] = create_user_vma(A->vm, SZ_64K);
+	if (IS_ERR(result[1])) {
+		err = PTR_ERR(result[1]);
+		goto err_result0;
+	}
+
+	rq = record_registers(A, result[0], result[1], sema);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_result1;
+	}
+
+	err = poison_registers(B, poison, sema);
+	if (err) {
+		WRITE_ONCE(*sema, -1);
+		i915_request_put(rq);
+		goto err_result1;
+	}
+
+	if (i915_request_wait(rq, 0, HZ / 2) < 0) {
+		i915_request_put(rq);
+		err = -ETIME;
+		goto err_result1;
+	}
+	i915_request_put(rq);
+
+	err = compare_isolation(engine, ref, result, A, poison);
+
+err_result1:
+	i915_vma_put(result[1]);
+err_result0:
+	i915_vma_put(result[0]);
+err_ref1:
+	i915_vma_put(ref[1]);
+err_ref0:
+	i915_vma_put(ref[0]);
+err_B:
+	intel_context_put(B);
+err_A:
+	intel_context_put(A);
+	return err;
+}
+
+static bool skip_isolation(const struct intel_engine_cs *engine)
+{
+	if (engine->class == COPY_ENGINE_CLASS && INTEL_GEN(engine->i915) == 9)
+		return true;
+
+	if (engine->class == RENDER_CLASS && INTEL_GEN(engine->i915) == 11)
+		return true;
+
+	return false;
+}
+
+static int live_lrc_isolation(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	const u32 poison[] = {
+		STACK_MAGIC,
+		0x3a3a3a3a,
+		0x5c5c5c5c,
+		0xffffffff,
+		0xffff0000,
+	};
+
+	/*
+	 * Our goal is try and verify that per-context state cannot be
+	 * tampered with by another non-privileged client.
+	 *
+	 * We take the list of context registers from the LRI in the default
+	 * context image and attempt to modify that list from a remote context.
+	 */
+
+	for_each_engine(engine, gt, id) {
+		int err = 0;
+		int i;
+
+		/* Just don't even ask */
+		if (!IS_ENABLED(CONFIG_DRM_I915_SELFTEST_BROKEN) &&
+		    skip_isolation(engine))
+			continue;
+
+		intel_engine_pm_get(engine);
+		if (engine->pinned_default_state) {
+			for (i = 0; i < ARRAY_SIZE(poison); i++) {
+				err = __lrc_isolation(engine, poison[i]);
+				if (err)
+					break;
+
+				err = __lrc_isolation(engine, ~poison[i]);
+				if (err)
+					break;
+			}
+		}
+		intel_engine_pm_put(engine);
+		if (igt_flush_test(gt->i915))
+			err = -EIO;
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+static void garbage_reset(struct intel_engine_cs *engine,
+			  struct i915_request *rq)
+{
+	const unsigned int bit = I915_RESET_ENGINE + engine->id;
+	unsigned long *lock = &engine->gt->reset.flags;
+
+	if (test_and_set_bit(bit, lock))
+		return;
+
+	tasklet_disable(&engine->execlists.tasklet);
+
+	if (!rq->fence.error)
+		intel_engine_reset(engine, NULL);
+
+	tasklet_enable(&engine->execlists.tasklet);
+	clear_and_wake_up_bit(bit, lock);
+}
+
+static struct i915_request *garbage(struct intel_context *ce,
+				    struct rnd_state *prng)
+{
+	struct i915_request *rq;
+	int err;
+
+	err = intel_context_pin(ce);
+	if (err)
+		return ERR_PTR(err);
+
+	prandom_bytes_state(prng,
+			    ce->lrc_reg_state,
+			    ce->engine->context_size -
+			    LRC_STATE_PN * PAGE_SIZE);
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_unpin;
+	}
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+	return rq;
+
+err_unpin:
+	intel_context_unpin(ce);
+	return ERR_PTR(err);
+}
+
+static int __lrc_garbage(struct intel_engine_cs *engine, struct rnd_state *prng)
+{
+	struct intel_context *ce;
+	struct i915_request *hang;
+	int err = 0;
+
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	hang = garbage(ce, prng);
+	if (IS_ERR(hang)) {
+		err = PTR_ERR(hang);
+		goto err_ce;
+	}
+
+	if (wait_for_submit(engine, hang, HZ / 2)) {
+		i915_request_put(hang);
+		err = -ETIME;
+		goto err_ce;
+	}
+
+	intel_context_set_banned(ce);
+	garbage_reset(engine, hang);
+
+	intel_engine_flush_submission(engine);
+	if (!hang->fence.error) {
+		i915_request_put(hang);
+		pr_err("%s: corrupted context was not reset\n",
+		       engine->name);
+		err = -EINVAL;
+		goto err_ce;
+	}
+
+	if (i915_request_wait(hang, 0, HZ / 2) < 0) {
+		pr_err("%s: corrupted context did not recover\n",
+		       engine->name);
+		i915_request_put(hang);
+		err = -EIO;
+		goto err_ce;
+	}
+	i915_request_put(hang);
+
+err_ce:
+	intel_context_put(ce);
+	return err;
+}
+
+static int live_lrc_garbage(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	/*
+	 * Verify that we can recover if one context state is completely
+	 * corrupted.
+	 */
+
+	if (!IS_ENABLED(CONFIG_DRM_I915_SELFTEST_BROKEN))
+		return 0;
+
+	for_each_engine(engine, gt, id) {
+		I915_RND_STATE(prng);
+		int err = 0, i;
+
+		if (!intel_has_reset_engine(engine->gt))
+			continue;
+
+		intel_engine_pm_get(engine);
+		for (i = 0; i < 3; i++) {
+			err = __lrc_garbage(engine, &prng);
+			if (err)
+				break;
+		}
+		intel_engine_pm_put(engine);
+
+		if (igt_flush_test(gt->i915))
+			err = -EIO;
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 static int __live_pphwsp_runtime(struct intel_engine_cs *engine)
 {
 	struct intel_context *ce;
@@ -4845,7 +5565,9 @@ int intel_lrc_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(live_lrc_fixed),
 		SUBTEST(live_lrc_state),
 		SUBTEST(live_lrc_gpr),
+		SUBTEST(live_lrc_isolation),
 		SUBTEST(live_lrc_timestamp),
+		SUBTEST(live_lrc_garbage),
 		SUBTEST(live_pphwsp_runtime),
 	};
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_ring_submission.c b/drivers/gpu/drm/i915/gt/selftest_ring_submission.c
new file mode 100644
index 000000000000..9995faadd7e8
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_ring_submission.c
@@ -0,0 +1,296 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2020 Intel Corporation
+ */
+
+#include "intel_engine_pm.h"
+#include "selftests/igt_flush_test.h"
+
+static struct i915_vma *create_wally(struct intel_engine_cs *engine)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	u32 *cs;
+	int err;
+
+	obj = i915_gem_object_create_internal(engine->i915, 4096);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	vma = i915_vma_instance(obj, engine->gt->vm, NULL);
+	if (IS_ERR(vma)) {
+		i915_gem_object_put(obj);
+		return vma;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_HIGH);
+	if (err) {
+		i915_gem_object_put(obj);
+		return ERR_PTR(err);
+	}
+
+	err = i915_vma_sync(vma);
+	if (err) {
+		i915_gem_object_put(obj);
+		return ERR_PTR(err);
+	}
+
+	cs = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	if (IS_ERR(cs)) {
+		i915_gem_object_put(obj);
+		return ERR_CAST(cs);
+	}
+
+	if (INTEL_GEN(engine->i915) >= 6) {
+		*cs++ = MI_STORE_DWORD_IMM_GEN4;
+		*cs++ = 0;
+	} else if (INTEL_GEN(engine->i915) >= 4) {
+		*cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
+		*cs++ = 0;
+	} else {
+		*cs++ = MI_STORE_DWORD_IMM | MI_MEM_VIRTUAL;
+	}
+	*cs++ = vma->node.start + 4000;
+	*cs++ = STACK_MAGIC;
+
+	*cs++ = MI_BATCH_BUFFER_END;
+	i915_gem_object_unpin_map(obj);
+
+	vma->private = intel_context_create(engine); /* dummy residuals */
+	if (IS_ERR(vma->private)) {
+		vma = ERR_CAST(vma->private);
+		i915_gem_object_put(obj);
+	}
+
+	return vma;
+}
+
+static int context_sync(struct intel_context *ce)
+{
+	struct i915_request *rq;
+	int err = 0;
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	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 new_context_sync(struct intel_engine_cs *engine)
+{
+	struct intel_context *ce;
+	int err;
+
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	err = context_sync(ce);
+	intel_context_put(ce);
+
+	return err;
+}
+
+static int mixed_contexts_sync(struct intel_engine_cs *engine, u32 *result)
+{
+	int pass;
+	int err;
+
+	for (pass = 0; pass < 2; pass++) {
+		WRITE_ONCE(*result, 0);
+		err = context_sync(engine->kernel_context);
+		if (err || READ_ONCE(*result)) {
+			if (!err) {
+				pr_err("pass[%d] wa_bb emitted for the kernel context\n",
+				       pass);
+				err = -EINVAL;
+			}
+			return err;
+		}
+
+		WRITE_ONCE(*result, 0);
+		err = new_context_sync(engine);
+		if (READ_ONCE(*result) != STACK_MAGIC) {
+			if (!err) {
+				pr_err("pass[%d] wa_bb *NOT* emitted after the kernel context\n",
+				       pass);
+				err = -EINVAL;
+			}
+			return err;
+		}
+
+		WRITE_ONCE(*result, 0);
+		err = new_context_sync(engine);
+		if (READ_ONCE(*result) != STACK_MAGIC) {
+			if (!err) {
+				pr_err("pass[%d] wa_bb *NOT* emitted for the user context switch\n",
+				       pass);
+				err = -EINVAL;
+			}
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+static int double_context_sync_00(struct intel_engine_cs *engine, u32 *result)
+{
+	struct intel_context *ce;
+	int err, i;
+
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	for (i = 0; i < 2; i++) {
+		WRITE_ONCE(*result, 0);
+		err = context_sync(ce);
+		if (err)
+			break;
+	}
+	intel_context_put(ce);
+	if (err)
+		return err;
+
+	if (READ_ONCE(*result)) {
+		pr_err("wa_bb emitted between the same user context\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int kernel_context_sync_00(struct intel_engine_cs *engine, u32 *result)
+{
+	struct intel_context *ce;
+	int err, i;
+
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	for (i = 0; i < 2; i++) {
+		WRITE_ONCE(*result, 0);
+		err = context_sync(ce);
+		if (err)
+			break;
+
+		err = context_sync(engine->kernel_context);
+		if (err)
+			break;
+	}
+	intel_context_put(ce);
+	if (err)
+		return err;
+
+	if (READ_ONCE(*result)) {
+		pr_err("wa_bb emitted between the same user context [with intervening kernel]\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int __live_ctx_switch_wa(struct intel_engine_cs *engine)
+{
+	struct i915_vma *bb;
+	u32 *result;
+	int err;
+
+	bb = create_wally(engine);
+	if (IS_ERR(bb))
+		return PTR_ERR(bb);
+
+	result = i915_gem_object_pin_map(bb->obj, I915_MAP_WC);
+	if (IS_ERR(result)) {
+		intel_context_put(bb->private);
+		i915_vma_unpin_and_release(&bb, 0);
+		return PTR_ERR(result);
+	}
+	result += 1000;
+
+	engine->wa_ctx.vma = bb;
+
+	err = mixed_contexts_sync(engine, result);
+	if (err)
+		goto out;
+
+	err = double_context_sync_00(engine, result);
+	if (err)
+		goto out;
+
+	err = kernel_context_sync_00(engine, result);
+	if (err)
+		goto out;
+
+out:
+	intel_context_put(engine->wa_ctx.vma->private);
+	i915_vma_unpin_and_release(&engine->wa_ctx.vma, I915_VMA_RELEASE_MAP);
+	return err;
+}
+
+static int live_ctx_switch_wa(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	/*
+	 * Exercise the inter-context wa batch.
+	 *
+	 * Between each user context we run a wa batch, and since it may
+	 * have implications for user visible state, we have to check that
+	 * we do actually execute it.
+	 *
+	 * The trick we use is to replace the normal wa batch with a custom
+	 * one that writes to a marker within it, and we can then look for
+	 * that marker to confirm if the batch was run when we expect it,
+	 * and equally important it was wasn't run when we don't!
+	 */
+
+	for_each_engine(engine, gt, id) {
+		struct i915_vma *saved_wa;
+		int err;
+
+		if (!intel_engine_can_store_dword(engine))
+			continue;
+
+		if (IS_GEN_RANGE(gt->i915, 4, 5))
+			continue; /* MI_STORE_DWORD is privileged! */
+
+		saved_wa = fetch_and_zero(&engine->wa_ctx.vma);
+
+		intel_engine_pm_get(engine);
+		err = __live_ctx_switch_wa(engine);
+		intel_engine_pm_put(engine);
+		if (igt_flush_test(gt->i915))
+			err = -EIO;
+
+		engine->wa_ctx.vma = saved_wa;
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
+int intel_ring_submission_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_ctx_switch_wa),
+	};
+
+	if (HAS_EXECLISTS(i915))
+		return 0;
+
+	return intel_gt_live_subtests(tests, &i915->gt);
+}
diff --git a/drivers/gpu/drm/i915/gt/sysfs_engines.c b/drivers/gpu/drm/i915/gt/sysfs_engines.c
new file mode 100644
index 000000000000..8f9b2f33dbaf
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/sysfs_engines.c
@@ -0,0 +1,445 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/kobject.h>
+#include <linux/sysfs.h>
+
+#include "i915_drv.h"
+#include "intel_engine.h"
+#include "intel_engine_heartbeat.h"
+#include "sysfs_engines.h"
+
+struct kobj_engine {
+	struct kobject base;
+	struct intel_engine_cs *engine;
+};
+
+static struct intel_engine_cs *kobj_to_engine(struct kobject *kobj)
+{
+	return container_of(kobj, struct kobj_engine, base)->engine;
+}
+
+static ssize_t
+name_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	return sprintf(buf, "%s\n", kobj_to_engine(kobj)->name);
+}
+
+static struct kobj_attribute name_attr =
+__ATTR(name, 0444, name_show, NULL);
+
+static ssize_t
+class_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	return sprintf(buf, "%d\n", kobj_to_engine(kobj)->uabi_class);
+}
+
+static struct kobj_attribute class_attr =
+__ATTR(class, 0444, class_show, NULL);
+
+static ssize_t
+inst_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	return sprintf(buf, "%d\n", kobj_to_engine(kobj)->uabi_instance);
+}
+
+static struct kobj_attribute inst_attr =
+__ATTR(instance, 0444, inst_show, NULL);
+
+static ssize_t
+mmio_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	return sprintf(buf, "0x%x\n", kobj_to_engine(kobj)->mmio_base);
+}
+
+static struct kobj_attribute mmio_attr =
+__ATTR(mmio_base, 0444, mmio_show, NULL);
+
+static const char * const vcs_caps[] = {
+	[ilog2(I915_VIDEO_CLASS_CAPABILITY_HEVC)] = "hevc",
+	[ilog2(I915_VIDEO_AND_ENHANCE_CLASS_CAPABILITY_SFC)] = "sfc",
+};
+
+static const char * const vecs_caps[] = {
+	[ilog2(I915_VIDEO_AND_ENHANCE_CLASS_CAPABILITY_SFC)] = "sfc",
+};
+
+static ssize_t repr_trim(char *buf, ssize_t len)
+{
+	/* Trim off the trailing space and replace with a newline */
+	if (len > PAGE_SIZE)
+		len = PAGE_SIZE;
+	if (len > 0)
+		buf[len - 1] = '\n';
+
+	return len;
+}
+
+static ssize_t
+__caps_show(struct intel_engine_cs *engine,
+	    u32 caps, char *buf, bool show_unknown)
+{
+	const char * const *repr;
+	int count, n;
+	ssize_t len;
+
+	BUILD_BUG_ON(!typecheck(typeof(caps), engine->uabi_capabilities));
+
+	switch (engine->class) {
+	case VIDEO_DECODE_CLASS:
+		repr = vcs_caps;
+		count = ARRAY_SIZE(vcs_caps);
+		break;
+
+	case VIDEO_ENHANCEMENT_CLASS:
+		repr = vecs_caps;
+		count = ARRAY_SIZE(vecs_caps);
+		break;
+
+	default:
+		repr = NULL;
+		count = 0;
+		break;
+	}
+	GEM_BUG_ON(count > BITS_PER_TYPE(typeof(caps)));
+
+	len = 0;
+	for_each_set_bit(n,
+			 (unsigned long *)&caps,
+			 show_unknown ? BITS_PER_TYPE(typeof(caps)) : count) {
+		if (n >= count || !repr[n]) {
+			if (GEM_WARN_ON(show_unknown))
+				len += snprintf(buf + len, PAGE_SIZE - len,
+						"[%x] ", n);
+		} else {
+			len += snprintf(buf + len, PAGE_SIZE - len,
+					"%s ", repr[n]);
+		}
+		if (GEM_WARN_ON(len >= PAGE_SIZE))
+			break;
+	}
+	return repr_trim(buf, len);
+}
+
+static ssize_t
+caps_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return __caps_show(engine, engine->uabi_capabilities, buf, true);
+}
+
+static struct kobj_attribute caps_attr =
+__ATTR(capabilities, 0444, caps_show, NULL);
+
+static ssize_t
+all_caps_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	return __caps_show(kobj_to_engine(kobj), -1, buf, false);
+}
+
+static struct kobj_attribute all_caps_attr =
+__ATTR(known_capabilities, 0444, all_caps_show, NULL);
+
+static ssize_t
+max_spin_store(struct kobject *kobj, struct kobj_attribute *attr,
+	       const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long duration;
+	int err;
+
+	/*
+	 * When waiting for a request, if is it currently being executed
+	 * on the GPU, we busywait for a short while before sleeping. The
+	 * premise is that most requests are short, and if it is already
+	 * executing then there is a good chance that it will complete
+	 * before we can setup the interrupt handler and go to sleep.
+	 * We try to offset the cost of going to sleep, by first spinning
+	 * on the request -- if it completed in less time than it would take
+	 * to go sleep, process the interrupt and return back to the client,
+	 * then we have saved the client some latency, albeit at the cost
+	 * of spinning on an expensive CPU core.
+	 *
+	 * While we try to avoid waiting at all for a request that is unlikely
+	 * to complete, deciding how long it is worth spinning is for is an
+	 * arbitrary decision: trading off power vs latency.
+	 */
+
+	err = kstrtoull(buf, 0, &duration);
+	if (err)
+		return err;
+
+	if (duration > jiffies_to_nsecs(2))
+		return -EINVAL;
+
+	WRITE_ONCE(engine->props.max_busywait_duration_ns, duration);
+
+	return count;
+}
+
+static ssize_t
+max_spin_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.max_busywait_duration_ns);
+}
+
+static struct kobj_attribute max_spin_attr =
+__ATTR(max_busywait_duration_ns, 0644, max_spin_show, max_spin_store);
+
+static ssize_t
+timeslice_store(struct kobject *kobj, struct kobj_attribute *attr,
+		const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long duration;
+	int err;
+
+	/*
+	 * Execlists uses a scheduling quantum (a timeslice) to alternate
+	 * execution between ready-to-run contexts of equal priority. This
+	 * ensures that all users (though only if they of equal importance)
+	 * have the opportunity to run and prevents livelocks where contexts
+	 * may have implicit ordering due to userspace semaphores.
+	 */
+
+	err = kstrtoull(buf, 0, &duration);
+	if (err)
+		return err;
+
+	if (duration > jiffies_to_msecs(MAX_SCHEDULE_TIMEOUT))
+		return -EINVAL;
+
+	WRITE_ONCE(engine->props.timeslice_duration_ms, duration);
+
+	if (execlists_active(&engine->execlists))
+		set_timer_ms(&engine->execlists.timer, duration);
+
+	return count;
+}
+
+static ssize_t
+timeslice_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.timeslice_duration_ms);
+}
+
+static struct kobj_attribute timeslice_duration_attr =
+__ATTR(timeslice_duration_ms, 0644, timeslice_show, timeslice_store);
+
+static ssize_t
+stop_store(struct kobject *kobj, struct kobj_attribute *attr,
+	   const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long duration;
+	int err;
+
+	/*
+	 * When we allow ourselves to sleep before a GPU reset after disabling
+	 * submission, even for a few milliseconds, gives an innocent context
+	 * the opportunity to clear the GPU before the reset occurs. However,
+	 * how long to sleep depends on the typical non-preemptible duration
+	 * (a similar problem to determining the ideal preempt-reset timeout
+	 * or even the heartbeat interval).
+	 */
+
+	err = kstrtoull(buf, 0, &duration);
+	if (err)
+		return err;
+
+	if (duration > jiffies_to_msecs(MAX_SCHEDULE_TIMEOUT))
+		return -EINVAL;
+
+	WRITE_ONCE(engine->props.stop_timeout_ms, duration);
+	return count;
+}
+
+static ssize_t
+stop_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.stop_timeout_ms);
+}
+
+static struct kobj_attribute stop_timeout_attr =
+__ATTR(stop_timeout_ms, 0644, stop_show, stop_store);
+
+static ssize_t
+preempt_timeout_store(struct kobject *kobj, struct kobj_attribute *attr,
+		      const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long timeout;
+	int err;
+
+	/*
+	 * After initialising a preemption request, we give the current
+	 * resident a small amount of time to vacate the GPU. The preemption
+	 * request is for a higher priority context and should be immediate to
+	 * maintain high quality of service (and avoid priority inversion).
+	 * However, the preemption granularity of the GPU can be quite coarse
+	 * and so we need a compromise.
+	 */
+
+	err = kstrtoull(buf, 0, &timeout);
+	if (err)
+		return err;
+
+	if (timeout > jiffies_to_msecs(MAX_SCHEDULE_TIMEOUT))
+		return -EINVAL;
+
+	WRITE_ONCE(engine->props.preempt_timeout_ms, timeout);
+
+	if (READ_ONCE(engine->execlists.pending[0]))
+		set_timer_ms(&engine->execlists.preempt, timeout);
+
+	return count;
+}
+
+static ssize_t
+preempt_timeout_show(struct kobject *kobj, struct kobj_attribute *attr,
+		     char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.preempt_timeout_ms);
+}
+
+static struct kobj_attribute preempt_timeout_attr =
+__ATTR(preempt_timeout_ms, 0644, preempt_timeout_show, preempt_timeout_store);
+
+static ssize_t
+heartbeat_store(struct kobject *kobj, struct kobj_attribute *attr,
+		const char *buf, size_t count)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+	unsigned long long delay;
+	int err;
+
+	/*
+	 * We monitor the health of the system via periodic heartbeat pulses.
+	 * The pulses also provide the opportunity to perform garbage
+	 * collection.  However, we interpret an incomplete pulse (a missed
+	 * heartbeat) as an indication that the system is no longer responsive,
+	 * i.e. hung, and perform an engine or full GPU reset. Given that the
+	 * preemption granularity can be very coarse on a system, the optimal
+	 * value for any workload is unknowable!
+	 */
+
+	err = kstrtoull(buf, 0, &delay);
+	if (err)
+		return err;
+
+	if (delay >= jiffies_to_msecs(MAX_SCHEDULE_TIMEOUT))
+		return -EINVAL;
+
+	err = intel_engine_set_heartbeat(engine, delay);
+	if (err)
+		return err;
+
+	return count;
+}
+
+static ssize_t
+heartbeat_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
+{
+	struct intel_engine_cs *engine = kobj_to_engine(kobj);
+
+	return sprintf(buf, "%lu\n", engine->props.heartbeat_interval_ms);
+}
+
+static struct kobj_attribute heartbeat_interval_attr =
+__ATTR(heartbeat_interval_ms, 0644, heartbeat_show, heartbeat_store);
+
+static void kobj_engine_release(struct kobject *kobj)
+{
+	kfree(kobj);
+}
+
+static struct kobj_type kobj_engine_type = {
+	.release = kobj_engine_release,
+	.sysfs_ops = &kobj_sysfs_ops
+};
+
+static struct kobject *
+kobj_engine(struct kobject *dir, struct intel_engine_cs *engine)
+{
+	struct kobj_engine *ke;
+
+	ke = kzalloc(sizeof(*ke), GFP_KERNEL);
+	if (!ke)
+		return NULL;
+
+	kobject_init(&ke->base, &kobj_engine_type);
+	ke->engine = engine;
+
+	if (kobject_add(&ke->base, dir, "%s", engine->name)) {
+		kobject_put(&ke->base);
+		return NULL;
+	}
+
+	/* xfer ownership to sysfs tree */
+	return &ke->base;
+}
+
+void intel_engines_add_sysfs(struct drm_i915_private *i915)
+{
+	static const struct attribute *files[] = {
+		&name_attr.attr,
+		&class_attr.attr,
+		&inst_attr.attr,
+		&mmio_attr.attr,
+		&caps_attr.attr,
+		&all_caps_attr.attr,
+		&max_spin_attr.attr,
+		&stop_timeout_attr.attr,
+#if CONFIG_DRM_I915_HEARTBEAT_INTERVAL
+		&heartbeat_interval_attr.attr,
+#endif
+		NULL
+	};
+
+	struct device *kdev = i915->drm.primary->kdev;
+	struct intel_engine_cs *engine;
+	struct kobject *dir;
+
+	dir = kobject_create_and_add("engine", &kdev->kobj);
+	if (!dir)
+		return;
+
+	for_each_uabi_engine(engine, i915) {
+		struct kobject *kobj;
+
+		kobj = kobj_engine(dir, engine);
+		if (!kobj)
+			goto err_engine;
+
+		if (sysfs_create_files(kobj, files))
+			goto err_object;
+
+		if (intel_engine_has_timeslices(engine) &&
+		    sysfs_create_file(kobj, &timeslice_duration_attr.attr))
+			goto err_engine;
+
+		if (intel_engine_has_preempt_reset(engine) &&
+		    sysfs_create_file(kobj, &preempt_timeout_attr.attr))
+			goto err_engine;
+
+		if (0) {
+err_object:
+			kobject_put(kobj);
+err_engine:
+			dev_err(kdev, "Failed to add sysfs engine '%s'\n",
+				engine->name);
+			break;
+		}
+	}
+}
diff --git a/drivers/gpu/drm/i915/gt/sysfs_engines.h b/drivers/gpu/drm/i915/gt/sysfs_engines.h
new file mode 100644
index 000000000000..9546fffe03a7
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/sysfs_engines.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef INTEL_ENGINE_SYSFS_H
+#define INTEL_ENGINE_SYSFS_H
+
+struct drm_i915_private;
+
+void intel_engines_add_sysfs(struct drm_i915_private *i915);
+
+#endif /* INTEL_ENGINE_SYSFS_H */
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 1beaa77f9bb6..fe7778c28d2d 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
@@ -456,9 +456,7 @@ static void guc_reset_cancel(struct intel_engine_cs *engine)
 
 	/* Mark all executing requests as skipped. */
 	list_for_each_entry(rq, &engine->active.requests, sched.link) {
-		if (!i915_request_signaled(rq))
-			dma_fence_set_error(&rq->fence, -EIO);
-
+		i915_request_set_error_once(rq, -EIO);
 		i915_request_mark_complete(rq);
 	}
 
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 5434c07aefa1..18c755203688 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
@@ -43,7 +43,7 @@ void intel_uc_fw_change_status(struct intel_uc_fw *uc_fw,
  * features.
  */
 #define INTEL_UC_FIRMWARE_DEFS(fw_def, guc_def, huc_def) \
-	fw_def(TIGERLAKE,   0, guc_def(tgl, 35, 2, 0), huc_def(tgl,  7, 0, 3)) \
+	fw_def(TIGERLAKE,   0, guc_def(tgl, 35, 2, 0), huc_def(tgl,  7, 0, 12)) \
 	fw_def(ELKHARTLAKE, 0, guc_def(ehl, 33, 0, 4), huc_def(ehl,  9, 0, 0)) \
 	fw_def(ICELAKE,     0, guc_def(icl, 33, 0, 0), huc_def(icl,  9, 0, 0)) \
 	fw_def(COFFEELAKE,  5, guc_def(cml, 33, 0, 0), huc_def(cml,  4, 0, 0)) \
diff --git a/drivers/gpu/drm/i915/gvt/aperture_gm.c b/drivers/gpu/drm/i915/gvt/aperture_gm.c
index 771420453f82..8b13f091cee2 100644
--- a/drivers/gpu/drm/i915/gvt/aperture_gm.c
+++ b/drivers/gpu/drm/i915/gvt/aperture_gm.c
@@ -41,7 +41,7 @@
 static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct intel_gt *gt = gvt->gt;
 	unsigned int flags;
 	u64 start, end, size;
 	struct drm_mm_node *node;
@@ -61,14 +61,14 @@ static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm)
 		flags = PIN_MAPPABLE;
 	}
 
-	mutex_lock(&dev_priv->ggtt.vm.mutex);
-	mmio_hw_access_pre(dev_priv);
-	ret = i915_gem_gtt_insert(&dev_priv->ggtt.vm, node,
+	mutex_lock(&gt->ggtt->vm.mutex);
+	mmio_hw_access_pre(gt);
+	ret = i915_gem_gtt_insert(&gt->ggtt->vm, node,
 				  size, I915_GTT_PAGE_SIZE,
 				  I915_COLOR_UNEVICTABLE,
 				  start, end, flags);
-	mmio_hw_access_post(dev_priv);
-	mutex_unlock(&dev_priv->ggtt.vm.mutex);
+	mmio_hw_access_post(gt);
+	mutex_unlock(&gt->ggtt->vm.mutex);
 	if (ret)
 		gvt_err("fail to alloc %s gm space from host\n",
 			high_gm ? "high" : "low");
@@ -79,7 +79,7 @@ static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm)
 static int alloc_vgpu_gm(struct intel_vgpu *vgpu)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct intel_gt *gt = gvt->gt;
 	int ret;
 
 	ret = alloc_gm(vgpu, false);
@@ -98,20 +98,21 @@ static int alloc_vgpu_gm(struct intel_vgpu *vgpu)
 
 	return 0;
 out_free_aperture:
-	mutex_lock(&dev_priv->ggtt.vm.mutex);
+	mutex_lock(&gt->ggtt->vm.mutex);
 	drm_mm_remove_node(&vgpu->gm.low_gm_node);
-	mutex_unlock(&dev_priv->ggtt.vm.mutex);
+	mutex_unlock(&gt->ggtt->vm.mutex);
 	return ret;
 }
 
 static void free_vgpu_gm(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct intel_gvt *gvt = vgpu->gvt;
+	struct intel_gt *gt = gvt->gt;
 
-	mutex_lock(&dev_priv->ggtt.vm.mutex);
+	mutex_lock(&gt->ggtt->vm.mutex);
 	drm_mm_remove_node(&vgpu->gm.low_gm_node);
 	drm_mm_remove_node(&vgpu->gm.high_gm_node);
-	mutex_unlock(&dev_priv->ggtt.vm.mutex);
+	mutex_unlock(&gt->ggtt->vm.mutex);
 }
 
 /**
@@ -128,28 +129,29 @@ void intel_vgpu_write_fence(struct intel_vgpu *vgpu,
 		u32 fence, u64 value)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct drm_i915_private *i915 = gvt->gt->i915;
+	struct intel_uncore *uncore = gvt->gt->uncore;
 	struct i915_fence_reg *reg;
 	i915_reg_t fence_reg_lo, fence_reg_hi;
 
-	assert_rpm_wakelock_held(&dev_priv->runtime_pm);
+	assert_rpm_wakelock_held(uncore->rpm);
 
-	if (WARN_ON(fence >= vgpu_fence_sz(vgpu)))
+	if (drm_WARN_ON(&i915->drm, fence >= vgpu_fence_sz(vgpu)))
 		return;
 
 	reg = vgpu->fence.regs[fence];
-	if (WARN_ON(!reg))
+	if (drm_WARN_ON(&i915->drm, !reg))
 		return;
 
 	fence_reg_lo = FENCE_REG_GEN6_LO(reg->id);
 	fence_reg_hi = FENCE_REG_GEN6_HI(reg->id);
 
-	I915_WRITE(fence_reg_lo, 0);
-	POSTING_READ(fence_reg_lo);
+	intel_uncore_write(uncore, fence_reg_lo, 0);
+	intel_uncore_posting_read(uncore, fence_reg_lo);
 
-	I915_WRITE(fence_reg_hi, upper_32_bits(value));
-	I915_WRITE(fence_reg_lo, lower_32_bits(value));
-	POSTING_READ(fence_reg_lo);
+	intel_uncore_write(uncore, fence_reg_hi, upper_32_bits(value));
+	intel_uncore_write(uncore, fence_reg_lo, lower_32_bits(value));
+	intel_uncore_posting_read(uncore, fence_reg_lo);
 }
 
 static void _clear_vgpu_fence(struct intel_vgpu *vgpu)
@@ -163,42 +165,43 @@ static void _clear_vgpu_fence(struct intel_vgpu *vgpu)
 static void free_vgpu_fence(struct intel_vgpu *vgpu)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct intel_uncore *uncore = gvt->gt->uncore;
 	struct i915_fence_reg *reg;
+	intel_wakeref_t wakeref;
 	u32 i;
 
-	if (WARN_ON(!vgpu_fence_sz(vgpu)))
+	if (drm_WARN_ON(&gvt->gt->i915->drm, !vgpu_fence_sz(vgpu)))
 		return;
 
-	intel_runtime_pm_get(&dev_priv->runtime_pm);
+	wakeref = intel_runtime_pm_get(uncore->rpm);
 
-	mutex_lock(&dev_priv->ggtt.vm.mutex);
+	mutex_lock(&gvt->gt->ggtt->vm.mutex);
 	_clear_vgpu_fence(vgpu);
 	for (i = 0; i < vgpu_fence_sz(vgpu); i++) {
 		reg = vgpu->fence.regs[i];
 		i915_unreserve_fence(reg);
 		vgpu->fence.regs[i] = NULL;
 	}
-	mutex_unlock(&dev_priv->ggtt.vm.mutex);
+	mutex_unlock(&gvt->gt->ggtt->vm.mutex);
 
-	intel_runtime_pm_put_unchecked(&dev_priv->runtime_pm);
+	intel_runtime_pm_put(uncore->rpm, wakeref);
 }
 
 static int alloc_vgpu_fence(struct intel_vgpu *vgpu)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
-	struct intel_runtime_pm *rpm = &dev_priv->runtime_pm;
+	struct intel_uncore *uncore = gvt->gt->uncore;
 	struct i915_fence_reg *reg;
+	intel_wakeref_t wakeref;
 	int i;
 
-	intel_runtime_pm_get(rpm);
+	wakeref = intel_runtime_pm_get(uncore->rpm);
 
 	/* Request fences from host */
-	mutex_lock(&dev_priv->ggtt.vm.mutex);
+	mutex_lock(&gvt->gt->ggtt->vm.mutex);
 
 	for (i = 0; i < vgpu_fence_sz(vgpu); i++) {
-		reg = i915_reserve_fence(&dev_priv->ggtt);
+		reg = i915_reserve_fence(gvt->gt->ggtt);
 		if (IS_ERR(reg))
 			goto out_free_fence;
 
@@ -207,9 +210,10 @@ static int alloc_vgpu_fence(struct intel_vgpu *vgpu)
 
 	_clear_vgpu_fence(vgpu);
 
-	mutex_unlock(&dev_priv->ggtt.vm.mutex);
-	intel_runtime_pm_put_unchecked(rpm);
+	mutex_unlock(&gvt->gt->ggtt->vm.mutex);
+	intel_runtime_pm_put(uncore->rpm, wakeref);
 	return 0;
+
 out_free_fence:
 	gvt_vgpu_err("Failed to alloc fences\n");
 	/* Return fences to host, if fail */
@@ -220,8 +224,8 @@ out_free_fence:
 		i915_unreserve_fence(reg);
 		vgpu->fence.regs[i] = NULL;
 	}
-	mutex_unlock(&dev_priv->ggtt.vm.mutex);
-	intel_runtime_pm_put_unchecked(rpm);
+	mutex_unlock(&gvt->gt->ggtt->vm.mutex);
+	intel_runtime_pm_put_unchecked(uncore->rpm);
 	return -ENOSPC;
 }
 
@@ -315,11 +319,11 @@ void intel_vgpu_free_resource(struct intel_vgpu *vgpu)
  */
 void intel_vgpu_reset_resource(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct intel_gvt *gvt = vgpu->gvt;
+	intel_wakeref_t wakeref;
 
-	intel_runtime_pm_get(&dev_priv->runtime_pm);
-	_clear_vgpu_fence(vgpu);
-	intel_runtime_pm_put_unchecked(&dev_priv->runtime_pm);
+	with_intel_runtime_pm(gvt->gt->uncore->rpm, wakeref)
+		_clear_vgpu_fence(vgpu);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/gvt/cfg_space.c b/drivers/gpu/drm/i915/gvt/cfg_space.c
index 19cf1bbe059d..072725a448db 100644
--- a/drivers/gpu/drm/i915/gvt/cfg_space.c
+++ b/drivers/gpu/drm/i915/gvt/cfg_space.c
@@ -106,10 +106,13 @@ static void vgpu_pci_cfg_mem_write(struct intel_vgpu *vgpu, unsigned int off,
 int intel_vgpu_emulate_cfg_read(struct intel_vgpu *vgpu, unsigned int offset,
 	void *p_data, unsigned int bytes)
 {
-	if (WARN_ON(bytes > 4))
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+
+	if (drm_WARN_ON(&i915->drm, bytes > 4))
 		return -EINVAL;
 
-	if (WARN_ON(offset + bytes > vgpu->gvt->device_info.cfg_space_size))
+	if (drm_WARN_ON(&i915->drm,
+			offset + bytes > vgpu->gvt->device_info.cfg_space_size))
 		return -EINVAL;
 
 	memcpy(p_data, vgpu_cfg_space(vgpu) + offset, bytes);
@@ -297,34 +300,36 @@ static int emulate_pci_bar_write(struct intel_vgpu *vgpu, unsigned int offset,
 int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset,
 	void *p_data, unsigned int bytes)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	int ret;
 
-	if (WARN_ON(bytes > 4))
+	if (drm_WARN_ON(&i915->drm, bytes > 4))
 		return -EINVAL;
 
-	if (WARN_ON(offset + bytes > vgpu->gvt->device_info.cfg_space_size))
+	if (drm_WARN_ON(&i915->drm,
+			offset + bytes > vgpu->gvt->device_info.cfg_space_size))
 		return -EINVAL;
 
 	/* First check if it's PCI_COMMAND */
 	if (IS_ALIGNED(offset, 2) && offset == PCI_COMMAND) {
-		if (WARN_ON(bytes > 2))
+		if (drm_WARN_ON(&i915->drm, bytes > 2))
 			return -EINVAL;
 		return emulate_pci_command_write(vgpu, offset, p_data, bytes);
 	}
 
 	switch (rounddown(offset, 4)) {
 	case PCI_ROM_ADDRESS:
-		if (WARN_ON(!IS_ALIGNED(offset, 4)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, 4)))
 			return -EINVAL;
 		return emulate_pci_rom_bar_write(vgpu, offset, p_data, bytes);
 
 	case PCI_BASE_ADDRESS_0 ... PCI_BASE_ADDRESS_5:
-		if (WARN_ON(!IS_ALIGNED(offset, 4)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, 4)))
 			return -EINVAL;
 		return emulate_pci_bar_write(vgpu, offset, p_data, bytes);
 
 	case INTEL_GVT_PCI_SWSCI:
-		if (WARN_ON(!IS_ALIGNED(offset, 4)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, 4)))
 			return -EINVAL;
 		ret = intel_vgpu_emulate_opregion_request(vgpu, *(u32 *)p_data);
 		if (ret)
@@ -332,7 +337,7 @@ int intel_vgpu_emulate_cfg_write(struct intel_vgpu *vgpu, unsigned int offset,
 		break;
 
 	case INTEL_GVT_PCI_OPREGION:
-		if (WARN_ON(!IS_ALIGNED(offset, 4)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, 4)))
 			return -EINVAL;
 		ret = intel_vgpu_opregion_base_write_handler(vgpu,
 						   *(u32 *)p_data);
@@ -391,9 +396,9 @@ void intel_vgpu_init_cfg_space(struct intel_vgpu *vgpu,
 	memset(vgpu_cfg_space(vgpu) + INTEL_GVT_PCI_OPREGION, 0, 4);
 
 	vgpu->cfg_space.bar[INTEL_GVT_PCI_BAR_GTTMMIO].size =
-				pci_resource_len(gvt->dev_priv->drm.pdev, 0);
+		pci_resource_len(gvt->gt->i915->drm.pdev, 0);
 	vgpu->cfg_space.bar[INTEL_GVT_PCI_BAR_APERTURE].size =
-				pci_resource_len(gvt->dev_priv->drm.pdev, 2);
+		pci_resource_len(gvt->gt->i915->drm.pdev, 2);
 
 	memset(vgpu_cfg_space(vgpu) + PCI_ROM_ADDRESS, 0, 4);
 }
diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index 21a176cd8acc..9e065ad0658f 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -462,7 +462,7 @@ enum {
 
 struct parser_exec_state {
 	struct intel_vgpu *vgpu;
-	int ring_id;
+	const struct intel_engine_cs *engine;
 
 	int buf_type;
 
@@ -635,39 +635,42 @@ static const struct decode_info *ring_decode_info[I915_NUM_ENGINES][8] = {
 	},
 };
 
-static inline u32 get_opcode(u32 cmd, int ring_id)
+static inline u32 get_opcode(u32 cmd, const struct intel_engine_cs *engine)
 {
 	const struct decode_info *d_info;
 
-	d_info = ring_decode_info[ring_id][CMD_TYPE(cmd)];
+	d_info = ring_decode_info[engine->id][CMD_TYPE(cmd)];
 	if (d_info == NULL)
 		return INVALID_OP;
 
 	return cmd >> (32 - d_info->op_len);
 }
 
-static inline const struct cmd_info *find_cmd_entry(struct intel_gvt *gvt,
-		unsigned int opcode, int ring_id)
+static inline const struct cmd_info *
+find_cmd_entry(struct intel_gvt *gvt, unsigned int opcode,
+	       const struct intel_engine_cs *engine)
 {
 	struct cmd_entry *e;
 
 	hash_for_each_possible(gvt->cmd_table, e, hlist, opcode) {
-		if (opcode == e->info->opcode && e->info->rings & BIT(ring_id))
+		if (opcode == e->info->opcode &&
+		    e->info->rings & engine->mask)
 			return e->info;
 	}
 	return NULL;
 }
 
-static inline const struct cmd_info *get_cmd_info(struct intel_gvt *gvt,
-		u32 cmd, int ring_id)
+static inline const struct cmd_info *
+get_cmd_info(struct intel_gvt *gvt, u32 cmd,
+	     const struct intel_engine_cs *engine)
 {
 	u32 opcode;
 
-	opcode = get_opcode(cmd, ring_id);
+	opcode = get_opcode(cmd, engine);
 	if (opcode == INVALID_OP)
 		return NULL;
 
-	return find_cmd_entry(gvt, opcode, ring_id);
+	return find_cmd_entry(gvt, opcode, engine);
 }
 
 static inline u32 sub_op_val(u32 cmd, u32 hi, u32 low)
@@ -675,12 +678,12 @@ static inline u32 sub_op_val(u32 cmd, u32 hi, u32 low)
 	return (cmd >> low) & ((1U << (hi - low + 1)) - 1);
 }
 
-static inline void print_opcode(u32 cmd, int ring_id)
+static inline void print_opcode(u32 cmd, const struct intel_engine_cs *engine)
 {
 	const struct decode_info *d_info;
 	int i;
 
-	d_info = ring_decode_info[ring_id][CMD_TYPE(cmd)];
+	d_info = ring_decode_info[engine->id][CMD_TYPE(cmd)];
 	if (d_info == NULL)
 		return;
 
@@ -709,10 +712,11 @@ static void parser_exec_state_dump(struct parser_exec_state *s)
 	int cnt = 0;
 	int i;
 
-	gvt_dbg_cmd("  vgpu%d RING%d: ring_start(%08lx) ring_end(%08lx)"
-			" ring_head(%08lx) ring_tail(%08lx)\n", s->vgpu->id,
-			s->ring_id, s->ring_start, s->ring_start + s->ring_size,
-			s->ring_head, s->ring_tail);
+	gvt_dbg_cmd("  vgpu%d RING%s: ring_start(%08lx) ring_end(%08lx)"
+		    " ring_head(%08lx) ring_tail(%08lx)\n",
+		    s->vgpu->id, s->engine->name,
+		    s->ring_start, s->ring_start + s->ring_size,
+		    s->ring_head, s->ring_tail);
 
 	gvt_dbg_cmd("  %s %s ip_gma(%08lx) ",
 			s->buf_type == RING_BUFFER_INSTRUCTION ?
@@ -729,7 +733,7 @@ static void parser_exec_state_dump(struct parser_exec_state *s)
 			s->ip_va, cmd_val(s, 0), cmd_val(s, 1),
 			cmd_val(s, 2), cmd_val(s, 3));
 
-	print_opcode(cmd_val(s, 0), s->ring_id);
+	print_opcode(cmd_val(s, 0), s->engine);
 
 	s->ip_va = (u32 *)((((u64)s->ip_va) >> 12) << 12);
 
@@ -840,7 +844,6 @@ static int force_nonpriv_reg_handler(struct parser_exec_state *s,
 	unsigned int data;
 	u32 ring_base;
 	u32 nopid;
-	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
 
 	if (!strcmp(cmd, "lri"))
 		data = cmd_val(s, index + 1);
@@ -850,7 +853,7 @@ static int force_nonpriv_reg_handler(struct parser_exec_state *s,
 		return -EINVAL;
 	}
 
-	ring_base = dev_priv->engine[s->ring_id]->mmio_base;
+	ring_base = s->engine->mmio_base;
 	nopid = i915_mmio_reg_offset(RING_NOPID(ring_base));
 
 	if (!intel_gvt_in_force_nonpriv_whitelist(gvt, data) &&
@@ -926,9 +929,9 @@ static int cmd_reg_handler(struct parser_exec_state *s,
 	 * update reg values in it into vregs, so LRIs in workload with
 	 * inhibit context will restore with correct values
 	 */
-	if (IS_GEN(gvt->dev_priv, 9) &&
-			intel_gvt_mmio_is_in_ctx(gvt, offset) &&
-			!strncmp(cmd, "lri", 3)) {
+	if (IS_GEN(s->engine->i915, 9) &&
+	    intel_gvt_mmio_is_in_ctx(gvt, offset) &&
+	    !strncmp(cmd, "lri", 3)) {
 		intel_gvt_hypervisor_read_gpa(s->vgpu,
 			s->workload->ring_context_gpa + 12, &ctx_sr_ctl, 4);
 		/* check inhibit context */
@@ -964,7 +967,6 @@ static int cmd_handler_lri(struct parser_exec_state *s)
 {
 	int i, ret = 0;
 	int cmd_len = cmd_length(s);
-	struct intel_gvt *gvt = s->vgpu->gvt;
 	u32 valid_len = CMD_LEN(1);
 
 	/*
@@ -979,8 +981,8 @@ static int cmd_handler_lri(struct parser_exec_state *s)
 	}
 
 	for (i = 1; i < cmd_len; i += 2) {
-		if (IS_BROADWELL(gvt->dev_priv) && s->ring_id != RCS0) {
-			if (s->ring_id == BCS0 &&
+		if (IS_BROADWELL(s->engine->i915) && s->engine->id != RCS0) {
+			if (s->engine->id == BCS0 &&
 			    cmd_reg(s, i) == i915_mmio_reg_offset(DERRMR))
 				ret |= 0;
 			else
@@ -1001,9 +1003,9 @@ static int cmd_handler_lrr(struct parser_exec_state *s)
 	int cmd_len = cmd_length(s);
 
 	for (i = 1; i < cmd_len; i += 2) {
-		if (IS_BROADWELL(s->vgpu->gvt->dev_priv))
+		if (IS_BROADWELL(s->engine->i915))
 			ret |= ((cmd_reg_inhibit(s, i) ||
-					(cmd_reg_inhibit(s, i + 1)))) ?
+				 (cmd_reg_inhibit(s, i + 1)))) ?
 				-EBADRQC : 0;
 		if (ret)
 			break;
@@ -1029,7 +1031,7 @@ static int cmd_handler_lrm(struct parser_exec_state *s)
 	int cmd_len = cmd_length(s);
 
 	for (i = 1; i < cmd_len;) {
-		if (IS_BROADWELL(gvt->dev_priv))
+		if (IS_BROADWELL(s->engine->i915))
 			ret |= (cmd_reg_inhibit(s, i)) ? -EBADRQC : 0;
 		if (ret)
 			break;
@@ -1141,7 +1143,7 @@ static int cmd_handler_pipe_control(struct parser_exec_state *s)
 				if (ret)
 					return ret;
 				if (index_mode) {
-					hws_pga = s->vgpu->hws_pga[s->ring_id];
+					hws_pga = s->vgpu->hws_pga[s->engine->id];
 					gma = hws_pga + gma;
 					patch_value(s, cmd_ptr(s, 2), gma);
 					val = cmd_val(s, 1) & (~(1 << 21));
@@ -1155,15 +1157,15 @@ static int cmd_handler_pipe_control(struct parser_exec_state *s)
 		return ret;
 
 	if (cmd_val(s, 1) & PIPE_CONTROL_NOTIFY)
-		set_bit(cmd_interrupt_events[s->ring_id].pipe_control_notify,
-				s->workload->pending_events);
+		set_bit(cmd_interrupt_events[s->engine->id].pipe_control_notify,
+			s->workload->pending_events);
 	return 0;
 }
 
 static int cmd_handler_mi_user_interrupt(struct parser_exec_state *s)
 {
-	set_bit(cmd_interrupt_events[s->ring_id].mi_user_interrupt,
-			s->workload->pending_events);
+	set_bit(cmd_interrupt_events[s->engine->id].mi_user_interrupt,
+		s->workload->pending_events);
 	patch_value(s, cmd_ptr(s, 0), MI_NOOP);
 	return 0;
 }
@@ -1213,7 +1215,7 @@ struct plane_code_mapping {
 static int gen8_decode_mi_display_flip(struct parser_exec_state *s,
 		struct mi_display_flip_command_info *info)
 {
-	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = s->engine->i915;
 	struct plane_code_mapping gen8_plane_code[] = {
 		[0] = {PIPE_A, PLANE_A, PRIMARY_A_FLIP_DONE},
 		[1] = {PIPE_B, PLANE_A, PRIMARY_B_FLIP_DONE},
@@ -1230,7 +1232,7 @@ static int gen8_decode_mi_display_flip(struct parser_exec_state *s,
 	dword2 = cmd_val(s, 2);
 
 	v = (dword0 & GENMASK(21, 19)) >> 19;
-	if (WARN_ON(v >= ARRAY_SIZE(gen8_plane_code)))
+	if (drm_WARN_ON(&dev_priv->drm, v >= ARRAY_SIZE(gen8_plane_code)))
 		return -EBADRQC;
 
 	info->pipe = gen8_plane_code[v].pipe;
@@ -1250,7 +1252,7 @@ static int gen8_decode_mi_display_flip(struct parser_exec_state *s,
 		info->stride_reg = SPRSTRIDE(info->pipe);
 		info->surf_reg = SPRSURF(info->pipe);
 	} else {
-		WARN_ON(1);
+		drm_WARN_ON(&dev_priv->drm, 1);
 		return -EBADRQC;
 	}
 	return 0;
@@ -1259,7 +1261,7 @@ static int gen8_decode_mi_display_flip(struct parser_exec_state *s,
 static int skl_decode_mi_display_flip(struct parser_exec_state *s,
 		struct mi_display_flip_command_info *info)
 {
-	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = s->engine->i915;
 	struct intel_vgpu *vgpu = s->vgpu;
 	u32 dword0 = cmd_val(s, 0);
 	u32 dword1 = cmd_val(s, 1);
@@ -1318,13 +1320,12 @@ static int skl_decode_mi_display_flip(struct parser_exec_state *s,
 static int gen8_check_mi_display_flip(struct parser_exec_state *s,
 		struct mi_display_flip_command_info *info)
 {
-	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
 	u32 stride, tile;
 
 	if (!info->async_flip)
 		return 0;
 
-	if (INTEL_GEN(dev_priv) >= 9) {
+	if (INTEL_GEN(s->engine->i915) >= 9) {
 		stride = vgpu_vreg_t(s->vgpu, info->stride_reg) & GENMASK(9, 0);
 		tile = (vgpu_vreg_t(s->vgpu, info->ctrl_reg) &
 				GENMASK(12, 10)) >> 10;
@@ -1347,7 +1348,7 @@ static int gen8_update_plane_mmio_from_mi_display_flip(
 		struct parser_exec_state *s,
 		struct mi_display_flip_command_info *info)
 {
-	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = s->engine->i915;
 	struct intel_vgpu *vgpu = s->vgpu;
 
 	set_mask_bits(&vgpu_vreg_t(vgpu, info->surf_reg), GENMASK(31, 12),
@@ -1378,11 +1379,9 @@ static int gen8_update_plane_mmio_from_mi_display_flip(
 static int decode_mi_display_flip(struct parser_exec_state *s,
 		struct mi_display_flip_command_info *info)
 {
-	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
-
-	if (IS_BROADWELL(dev_priv))
+	if (IS_BROADWELL(s->engine->i915))
 		return gen8_decode_mi_display_flip(s, info);
-	if (INTEL_GEN(dev_priv) >= 9)
+	if (INTEL_GEN(s->engine->i915) >= 9)
 		return skl_decode_mi_display_flip(s, info);
 
 	return -ENODEV;
@@ -1667,7 +1666,7 @@ static int cmd_handler_mi_flush_dw(struct parser_exec_state *s)
 		if (ret)
 			return ret;
 		if (index_mode) {
-			hws_pga = s->vgpu->hws_pga[s->ring_id];
+			hws_pga = s->vgpu->hws_pga[s->engine->id];
 			gma = hws_pga + gma;
 			patch_value(s, cmd_ptr(s, 1), gma);
 			val = cmd_val(s, 0) & (~(1 << 21));
@@ -1676,8 +1675,8 @@ static int cmd_handler_mi_flush_dw(struct parser_exec_state *s)
 	}
 	/* Check notify bit */
 	if ((cmd_val(s, 0) & (1 << 8)))
-		set_bit(cmd_interrupt_events[s->ring_id].mi_flush_dw,
-				s->workload->pending_events);
+		set_bit(cmd_interrupt_events[s->engine->id].mi_flush_dw,
+			s->workload->pending_events);
 	return ret;
 }
 
@@ -1725,12 +1724,18 @@ static int copy_gma_to_hva(struct intel_vgpu *vgpu, struct intel_vgpu_mm *mm,
 static int batch_buffer_needs_scan(struct parser_exec_state *s)
 {
 	/* Decide privilege based on address space */
-	if (cmd_val(s, 0) & (1 << 8) &&
-			!(s->vgpu->scan_nonprivbb & (1 << s->ring_id)))
+	if (cmd_val(s, 0) & BIT(8) &&
+	    !(s->vgpu->scan_nonprivbb & s->engine->mask))
 		return 0;
+
 	return 1;
 }
 
+static const char *repr_addr_type(unsigned int type)
+{
+	return type == PPGTT_BUFFER ? "ppgtt" : "ggtt";
+}
+
 static int find_bb_size(struct parser_exec_state *s,
 			unsigned long *bb_size,
 			unsigned long *bb_end_cmd_offset)
@@ -1753,24 +1758,24 @@ static int find_bb_size(struct parser_exec_state *s,
 		return -EFAULT;
 
 	cmd = cmd_val(s, 0);
-	info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
+	info = get_cmd_info(s->vgpu->gvt, cmd, s->engine);
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
-				cmd, get_opcode(cmd, s->ring_id),
-				(s->buf_addr_type == PPGTT_BUFFER) ?
-				"ppgtt" : "ggtt", s->ring_id, s->workload);
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %s, workload=%p\n",
+			     cmd, get_opcode(cmd, s->engine),
+			     repr_addr_type(s->buf_addr_type),
+			     s->engine->name, s->workload);
 		return -EBADRQC;
 	}
 	do {
 		if (copy_gma_to_hva(s->vgpu, mm,
-				gma, gma + 4, &cmd) < 0)
+				    gma, gma + 4, &cmd) < 0)
 			return -EFAULT;
-		info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
+		info = get_cmd_info(s->vgpu->gvt, cmd, s->engine);
 		if (info == NULL) {
-			gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
-				cmd, get_opcode(cmd, s->ring_id),
-				(s->buf_addr_type == PPGTT_BUFFER) ?
-				"ppgtt" : "ggtt", s->ring_id, s->workload);
+			gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %s, workload=%p\n",
+				     cmd, get_opcode(cmd, s->engine),
+				     repr_addr_type(s->buf_addr_type),
+				     s->engine->name, s->workload);
 			return -EBADRQC;
 		}
 
@@ -1799,12 +1804,12 @@ static int audit_bb_end(struct parser_exec_state *s, void *va)
 	u32 cmd = *(u32 *)va;
 	const struct cmd_info *info;
 
-	info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
+	info = get_cmd_info(s->vgpu->gvt, cmd, s->engine);
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
-			cmd, get_opcode(cmd, s->ring_id),
-			(s->buf_addr_type == PPGTT_BUFFER) ?
-			"ppgtt" : "ggtt", s->ring_id, s->workload);
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %s, workload=%p\n",
+			     cmd, get_opcode(cmd, s->engine),
+			     repr_addr_type(s->buf_addr_type),
+			     s->engine->name, s->workload);
 		return -EBADRQC;
 	}
 
@@ -1857,7 +1862,7 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	if (bb->ppgtt)
 		start_offset = gma & ~I915_GTT_PAGE_MASK;
 
-	bb->obj = i915_gem_object_create_shmem(s->vgpu->gvt->dev_priv,
+	bb->obj = i915_gem_object_create_shmem(s->engine->i915,
 					       round_up(bb_size + start_offset,
 							PAGE_SIZE));
 	if (IS_ERR(bb->obj)) {
@@ -2666,25 +2671,25 @@ static int cmd_parser_exec(struct parser_exec_state *s)
 	if (cmd == MI_NOOP)
 		info = &cmd_info[mi_noop_index];
 	else
-		info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
+		info = get_cmd_info(s->vgpu->gvt, cmd, s->engine);
 
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
-				cmd, get_opcode(cmd, s->ring_id),
-				(s->buf_addr_type == PPGTT_BUFFER) ?
-				"ppgtt" : "ggtt", s->ring_id, s->workload);
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %s, workload=%p\n",
+			     cmd, get_opcode(cmd, s->engine),
+			     repr_addr_type(s->buf_addr_type),
+			     s->engine->name, s->workload);
 		return -EBADRQC;
 	}
 
 	s->info = info;
 
-	trace_gvt_command(vgpu->id, s->ring_id, s->ip_gma, s->ip_va,
+	trace_gvt_command(vgpu->id, s->engine->id, s->ip_gma, s->ip_va,
 			  cmd_length(s), s->buf_type, s->buf_addr_type,
 			  s->workload, info->name);
 
 	if ((info->flag & F_LEN_MASK) == F_LEN_VAR_FIXED) {
 		ret = gvt_check_valid_cmd_length(cmd_length(s),
-			info->valid_len);
+						 info->valid_len);
 		if (ret)
 			return ret;
 	}
@@ -2781,7 +2786,7 @@ static int scan_workload(struct intel_vgpu_workload *workload)
 	s.buf_type = RING_BUFFER_INSTRUCTION;
 	s.buf_addr_type = GTT_BUFFER;
 	s.vgpu = workload->vgpu;
-	s.ring_id = workload->ring_id;
+	s.engine = workload->engine;
 	s.ring_start = workload->rb_start;
 	s.ring_size = _RING_CTL_BUF_SIZE(workload->rb_ctl);
 	s.ring_head = gma_head;
@@ -2790,8 +2795,7 @@ static int scan_workload(struct intel_vgpu_workload *workload)
 	s.workload = workload;
 	s.is_ctx_wa = false;
 
-	if ((bypass_scan_mask & (1 << workload->ring_id)) ||
-		gma_head == gma_tail)
+	if (bypass_scan_mask & workload->engine->mask || gma_head == gma_tail)
 		return 0;
 
 	ret = ip_gma_set(&s, gma_head);
@@ -2830,7 +2834,7 @@ static int scan_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	s.buf_type = RING_BUFFER_INSTRUCTION;
 	s.buf_addr_type = GTT_BUFFER;
 	s.vgpu = workload->vgpu;
-	s.ring_id = workload->ring_id;
+	s.engine = workload->engine;
 	s.ring_start = wa_ctx->indirect_ctx.guest_gma;
 	s.ring_size = ring_size;
 	s.ring_head = gma_head;
@@ -2855,7 +2859,6 @@ static int shadow_workload_ring_buffer(struct intel_vgpu_workload *workload)
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	unsigned long gma_head, gma_tail, gma_top, guest_rb_size;
 	void *shadow_ring_buffer_va;
-	int ring_id = workload->ring_id;
 	int ret;
 
 	guest_rb_size = _RING_CTL_BUF_SIZE(workload->rb_ctl);
@@ -2868,21 +2871,21 @@ static int shadow_workload_ring_buffer(struct intel_vgpu_workload *workload)
 	gma_tail = workload->rb_start + workload->rb_tail;
 	gma_top = workload->rb_start + guest_rb_size;
 
-	if (workload->rb_len > s->ring_scan_buffer_size[ring_id]) {
+	if (workload->rb_len > s->ring_scan_buffer_size[workload->engine->id]) {
 		void *p;
 
 		/* realloc the new ring buffer if needed */
-		p = krealloc(s->ring_scan_buffer[ring_id], workload->rb_len,
-				GFP_KERNEL);
+		p = krealloc(s->ring_scan_buffer[workload->engine->id],
+			     workload->rb_len, GFP_KERNEL);
 		if (!p) {
 			gvt_vgpu_err("fail to re-alloc ring scan buffer\n");
 			return -ENOMEM;
 		}
-		s->ring_scan_buffer[ring_id] = p;
-		s->ring_scan_buffer_size[ring_id] = workload->rb_len;
+		s->ring_scan_buffer[workload->engine->id] = p;
+		s->ring_scan_buffer_size[workload->engine->id] = workload->rb_len;
 	}
 
-	shadow_ring_buffer_va = s->ring_scan_buffer[ring_id];
+	shadow_ring_buffer_va = s->ring_scan_buffer[workload->engine->id];
 
 	/* get shadow ring buffer va */
 	workload->shadow_ring_buffer_va = shadow_ring_buffer_va;
@@ -2940,7 +2943,7 @@ static int shadow_indirect_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	int ret = 0;
 	void *map;
 
-	obj = i915_gem_object_create_shmem(workload->vgpu->gvt->dev_priv,
+	obj = i915_gem_object_create_shmem(workload->engine->i915,
 					   roundup(ctx_size + CACHELINE_BYTES,
 						   PAGE_SIZE));
 	if (IS_ERR(obj))
@@ -3029,30 +3032,14 @@ int intel_gvt_scan_and_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	return 0;
 }
 
-static const struct cmd_info *find_cmd_entry_any_ring(struct intel_gvt *gvt,
-		unsigned int opcode, unsigned long rings)
-{
-	const struct cmd_info *info = NULL;
-	unsigned int ring;
-
-	for_each_set_bit(ring, &rings, I915_NUM_ENGINES) {
-		info = find_cmd_entry(gvt, opcode, ring);
-		if (info)
-			break;
-	}
-	return info;
-}
-
 static int init_cmd_table(struct intel_gvt *gvt)
 {
+	unsigned int gen_type = intel_gvt_get_device_type(gvt);
 	int i;
-	struct cmd_entry *e;
-	const struct cmd_info *info;
-	unsigned int gen_type;
-
-	gen_type = intel_gvt_get_device_type(gvt);
 
 	for (i = 0; i < ARRAY_SIZE(cmd_info); i++) {
+		struct cmd_entry *e;
+
 		if (!(cmd_info[i].devices & gen_type))
 			continue;
 
@@ -3061,23 +3048,16 @@ static int init_cmd_table(struct intel_gvt *gvt)
 			return -ENOMEM;
 
 		e->info = &cmd_info[i];
-		info = find_cmd_entry_any_ring(gvt,
-				e->info->opcode, e->info->rings);
-		if (info) {
-			gvt_err("%s %s duplicated\n", e->info->name,
-					info->name);
-			kfree(e);
-			return -EEXIST;
-		}
 		if (cmd_info[i].opcode == OP_MI_NOOP)
 			mi_noop_index = i;
 
 		INIT_HLIST_NODE(&e->hlist);
 		add_cmd_entry(gvt, e);
 		gvt_dbg_cmd("add %-30s op %04x flag %x devs %02x rings %02x\n",
-				e->info->name, e->info->opcode, e->info->flag,
-				e->info->devices, e->info->rings);
+			    e->info->name, e->info->opcode, e->info->flag,
+			    e->info->devices, e->info->rings);
 	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/gvt/debugfs.c b/drivers/gpu/drm/i915/gvt/debugfs.c
index 285f6011a537..ec47d4114554 100644
--- a/drivers/gpu/drm/i915/gvt/debugfs.c
+++ b/drivers/gpu/drm/i915/gvt/debugfs.c
@@ -58,12 +58,11 @@ static int mmio_offset_compare(void *priv,
 static inline int mmio_diff_handler(struct intel_gvt *gvt,
 				    u32 offset, void *data)
 {
-	struct drm_i915_private *i915 = gvt->dev_priv;
 	struct mmio_diff_param *param = data;
 	struct diff_mmio *node;
 	u32 preg, vreg;
 
-	preg = intel_uncore_read_notrace(&i915->uncore, _MMIO(offset));
+	preg = intel_uncore_read_notrace(gvt->gt->uncore, _MMIO(offset));
 	vreg = vgpu_vreg(param->vgpu, offset);
 
 	if (preg != vreg) {
@@ -98,10 +97,10 @@ static int vgpu_mmio_diff_show(struct seq_file *s, void *unused)
 	mutex_lock(&gvt->lock);
 	spin_lock_bh(&gvt->scheduler.mmio_context_lock);
 
-	mmio_hw_access_pre(gvt->dev_priv);
+	mmio_hw_access_pre(gvt->gt);
 	/* Recognize all the diff mmios to list. */
 	intel_gvt_for_each_tracked_mmio(gvt, mmio_diff_handler, &param);
-	mmio_hw_access_post(gvt->dev_priv);
+	mmio_hw_access_post(gvt->gt);
 
 	spin_unlock_bh(&gvt->scheduler.mmio_context_lock);
 	mutex_unlock(&gvt->lock);
@@ -128,6 +127,7 @@ static int
 vgpu_scan_nonprivbb_get(void *data, u64 *val)
 {
 	struct intel_vgpu *vgpu = (struct intel_vgpu *)data;
+
 	*val = vgpu->scan_nonprivbb;
 	return 0;
 }
@@ -142,42 +142,7 @@ static int
 vgpu_scan_nonprivbb_set(void *data, u64 val)
 {
 	struct intel_vgpu *vgpu = (struct intel_vgpu *)data;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
-	enum intel_engine_id id;
-	char buf[128], *s;
-	int len;
-
-	val &= (1 << I915_NUM_ENGINES) - 1;
-
-	if (vgpu->scan_nonprivbb == val)
-		return 0;
-
-	if (!val)
-		goto done;
-
-	len = sprintf(buf,
-		"gvt: vgpu %d turns on non-privileged batch buffers scanning on Engines:",
-		vgpu->id);
-
-	s = buf + len;
-
-	for (id = 0; id < I915_NUM_ENGINES; id++) {
-		struct intel_engine_cs *engine;
-
-		engine = dev_priv->engine[id];
-		if (engine && (val & (1 << id))) {
-			len = snprintf(s, 4, "%d, ", engine->id);
-			s += len;
-		} else
-			val &=  ~(1 << id);
-	}
-
-	if (val)
-		sprintf(s, "low performance expected.");
-
-	pr_warn("%s\n", buf);
 
-done:
 	vgpu->scan_nonprivbb = val;
 	return 0;
 }
@@ -220,7 +185,7 @@ void intel_gvt_debugfs_remove_vgpu(struct intel_vgpu *vgpu)
  */
 void intel_gvt_debugfs_init(struct intel_gvt *gvt)
 {
-	struct drm_minor *minor = gvt->dev_priv->drm.primary;
+	struct drm_minor *minor = gvt->gt->i915->drm.primary;
 
 	gvt->debugfs_root = debugfs_create_dir("gvt", minor->debugfs_root);
 
diff --git a/drivers/gpu/drm/i915/gvt/display.c b/drivers/gpu/drm/i915/gvt/display.c
index e1c313da6c00..6e5c9885d9fe 100644
--- a/drivers/gpu/drm/i915/gvt/display.c
+++ b/drivers/gpu/drm/i915/gvt/display.c
@@ -57,7 +57,7 @@ static int get_edp_pipe(struct intel_vgpu *vgpu)
 
 static int edp_pipe_is_enabled(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 
 	if (!(vgpu_vreg_t(vgpu, PIPECONF(_PIPE_EDP)) & PIPECONF_ENABLE))
 		return 0;
@@ -69,9 +69,10 @@ static int edp_pipe_is_enabled(struct intel_vgpu *vgpu)
 
 int pipe_is_enabled(struct intel_vgpu *vgpu, int pipe)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 
-	if (WARN_ON(pipe < PIPE_A || pipe >= I915_MAX_PIPES))
+	if (drm_WARN_ON(&dev_priv->drm,
+			pipe < PIPE_A || pipe >= I915_MAX_PIPES))
 		return -EINVAL;
 
 	if (vgpu_vreg_t(vgpu, PIPECONF(pipe)) & PIPECONF_ENABLE)
@@ -168,7 +169,7 @@ static u8 dpcd_fix_data[DPCD_HEADER_SIZE] = {
 
 static void emulate_monitor_status_change(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	int pipe;
 
 	if (IS_BROXTON(dev_priv)) {
@@ -319,9 +320,10 @@ static void clean_virtual_dp_monitor(struct intel_vgpu *vgpu, int port_num)
 static int setup_virtual_dp_monitor(struct intel_vgpu *vgpu, int port_num,
 				    int type, unsigned int resolution)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_vgpu_port *port = intel_vgpu_port(vgpu, port_num);
 
-	if (WARN_ON(resolution >= GVT_EDID_NUM))
+	if (drm_WARN_ON(&i915->drm, resolution >= GVT_EDID_NUM))
 		return -EINVAL;
 
 	port->edid = kzalloc(sizeof(*(port->edid)), GFP_KERNEL);
@@ -389,7 +391,7 @@ void intel_gvt_check_vblank_emulation(struct intel_gvt *gvt)
 
 static void emulate_vblank_on_pipe(struct intel_vgpu *vgpu, int pipe)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	struct intel_vgpu_irq *irq = &vgpu->irq;
 	int vblank_event[] = {
 		[PIPE_A] = PIPE_A_VBLANK,
@@ -421,7 +423,7 @@ static void emulate_vblank(struct intel_vgpu *vgpu)
 	int pipe;
 
 	mutex_lock(&vgpu->vgpu_lock);
-	for_each_pipe(vgpu->gvt->dev_priv, pipe)
+	for_each_pipe(vgpu->gvt->gt->i915, pipe)
 		emulate_vblank_on_pipe(vgpu, pipe);
 	mutex_unlock(&vgpu->vgpu_lock);
 }
@@ -454,10 +456,11 @@ void intel_gvt_emulate_vblank(struct intel_gvt *gvt)
  */
 void intel_vgpu_emulate_hotplug(struct intel_vgpu *vgpu, bool connected)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 
 	/* TODO: add more platforms support */
-	if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
+	if (IS_SKYLAKE(i915) || IS_KABYLAKE(i915) ||
+	    IS_COFFEELAKE(i915)) {
 		if (connected) {
 			vgpu_vreg_t(vgpu, SFUSE_STRAP) |=
 				SFUSE_STRAP_DDID_DETECTED;
@@ -483,7 +486,7 @@ void intel_vgpu_emulate_hotplug(struct intel_vgpu *vgpu, bool connected)
  */
 void intel_vgpu_clean_display(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 
 	if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ||
 	    IS_COFFEELAKE(dev_priv))
@@ -505,7 +508,7 @@ void intel_vgpu_clean_display(struct intel_vgpu *vgpu)
  */
 int intel_vgpu_init_display(struct intel_vgpu *vgpu, u64 resolution)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 
 	intel_vgpu_init_i2c_edid(vgpu);
 
diff --git a/drivers/gpu/drm/i915/gvt/dmabuf.c b/drivers/gpu/drm/i915/gvt/dmabuf.c
index ae139f0877ae..37fc460414a8 100644
--- a/drivers/gpu/drm/i915/gvt/dmabuf.c
+++ b/drivers/gpu/drm/i915/gvt/dmabuf.c
@@ -67,11 +67,11 @@ static int vgpu_gem_get_pages(
 	u32 page_num;
 
 	fb_info = (struct intel_vgpu_fb_info *)obj->gvt_info;
-	if (WARN_ON(!fb_info))
+	if (drm_WARN_ON(&dev_priv->drm, !fb_info))
 		return -ENODEV;
 
 	vgpu = fb_info->obj->vgpu;
-	if (WARN_ON(!vgpu))
+	if (drm_WARN_ON(&dev_priv->drm, !vgpu))
 		return -ENODEV;
 
 	st = kmalloc(sizeof(*st), GFP_KERNEL);
@@ -417,7 +417,7 @@ static void update_fb_info(struct vfio_device_gfx_plane_info *gvt_dmabuf,
 
 int intel_vgpu_query_plane(struct intel_vgpu *vgpu, void *args)
 {
-	struct drm_device *dev = &vgpu->gvt->dev_priv->drm;
+	struct drm_device *dev = &vgpu->gvt->gt->i915->drm;
 	struct vfio_device_gfx_plane_info *gfx_plane_info = args;
 	struct intel_vgpu_dmabuf_obj *dmabuf_obj;
 	struct intel_vgpu_fb_info fb_info;
@@ -523,7 +523,7 @@ out:
 /* To associate an exposed dmabuf with the dmabuf_obj */
 int intel_vgpu_get_dmabuf(struct intel_vgpu *vgpu, unsigned int dmabuf_id)
 {
-	struct drm_device *dev = &vgpu->gvt->dev_priv->drm;
+	struct drm_device *dev = &vgpu->gvt->gt->i915->drm;
 	struct intel_vgpu_dmabuf_obj *dmabuf_obj;
 	struct drm_i915_gem_object *obj;
 	struct dma_buf *dmabuf;
diff --git a/drivers/gpu/drm/i915/gvt/edid.c b/drivers/gpu/drm/i915/gvt/edid.c
index 1fe6124918f1..190651df5db1 100644
--- a/drivers/gpu/drm/i915/gvt/edid.c
+++ b/drivers/gpu/drm/i915/gvt/edid.c
@@ -135,7 +135,7 @@ static void reset_gmbus_controller(struct intel_vgpu *vgpu)
 static int gmbus0_mmio_write(struct intel_vgpu *vgpu,
 			unsigned int offset, void *p_data, unsigned int bytes)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	int port, pin_select;
 
 	memcpy(&vgpu_vreg(vgpu, offset), p_data, bytes);
@@ -147,13 +147,13 @@ static int gmbus0_mmio_write(struct intel_vgpu *vgpu,
 	if (pin_select == 0)
 		return 0;
 
-	if (IS_BROXTON(dev_priv))
+	if (IS_BROXTON(i915))
 		port = bxt_get_port_from_gmbus0(pin_select);
-	else if (IS_COFFEELAKE(dev_priv))
+	else if (IS_COFFEELAKE(i915))
 		port = cnp_get_port_from_gmbus0(pin_select);
 	else
 		port = get_port_from_gmbus0(pin_select);
-	if (WARN_ON(port < 0))
+	if (drm_WARN_ON(&i915->drm, port < 0))
 		return 0;
 
 	vgpu->display.i2c_edid.state = I2C_GMBUS;
@@ -276,7 +276,9 @@ static int gmbus1_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 static int gmbus3_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 	void *p_data, unsigned int bytes)
 {
-	WARN_ON(1);
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+
+	drm_WARN_ON(&i915->drm, 1);
 	return 0;
 }
 
@@ -371,7 +373,9 @@ static int gmbus2_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 int intel_gvt_i2c_handle_gmbus_read(struct intel_vgpu *vgpu,
 	unsigned int offset, void *p_data, unsigned int bytes)
 {
-	if (WARN_ON(bytes > 8 && (offset & (bytes - 1))))
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+
+	if (drm_WARN_ON(&i915->drm, bytes > 8 && (offset & (bytes - 1))))
 		return -EINVAL;
 
 	if (offset == i915_mmio_reg_offset(PCH_GMBUS2))
@@ -399,7 +403,9 @@ int intel_gvt_i2c_handle_gmbus_read(struct intel_vgpu *vgpu,
 int intel_gvt_i2c_handle_gmbus_write(struct intel_vgpu *vgpu,
 		unsigned int offset, void *p_data, unsigned int bytes)
 {
-	if (WARN_ON(bytes > 8 && (offset & (bytes - 1))))
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+
+	if (drm_WARN_ON(&i915->drm, bytes > 8 && (offset & (bytes - 1))))
 		return -EINVAL;
 
 	if (offset == i915_mmio_reg_offset(PCH_GMBUS0))
@@ -473,6 +479,7 @@ void intel_gvt_i2c_handle_aux_ch_write(struct intel_vgpu *vgpu,
 				unsigned int offset,
 				void *p_data)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_vgpu_i2c_edid *i2c_edid = &vgpu->display.i2c_edid;
 	int msg_length, ret_msg_size;
 	int msg, addr, ctrl, op;
@@ -532,9 +539,9 @@ void intel_gvt_i2c_handle_aux_ch_write(struct intel_vgpu *vgpu,
 		 * support the gfx driver to do EDID access.
 		 */
 	} else {
-		if (WARN_ON((op & 0x1) != GVT_AUX_I2C_READ))
+		if (drm_WARN_ON(&i915->drm, (op & 0x1) != GVT_AUX_I2C_READ))
 			return;
-		if (WARN_ON(msg_length != 4))
+		if (drm_WARN_ON(&i915->drm, msg_length != 4))
 			return;
 		if (i2c_edid->edid_available && i2c_edid->slave_selected) {
 			unsigned char val = edid_get_byte(vgpu);
diff --git a/drivers/gpu/drm/i915/gvt/execlist.c b/drivers/gpu/drm/i915/gvt/execlist.c
index d6e7a1189bad..dd25c3024370 100644
--- a/drivers/gpu/drm/i915/gvt/execlist.c
+++ b/drivers/gpu/drm/i915/gvt/execlist.c
@@ -39,8 +39,7 @@
 #define _EL_OFFSET_STATUS_BUF   0x370
 #define _EL_OFFSET_STATUS_PTR   0x3A0
 
-#define execlist_ring_mmio(gvt, ring_id, offset) \
-	(gvt->dev_priv->engine[ring_id]->mmio_base + (offset))
+#define execlist_ring_mmio(e, offset) ((e)->mmio_base + (offset))
 
 #define valid_context(ctx) ((ctx)->valid)
 #define same_context(a, b) (((a)->context_id == (b)->context_id) && \
@@ -54,12 +53,12 @@ static int context_switch_events[] = {
 	[VECS0] = VECS_AS_CONTEXT_SWITCH,
 };
 
-static int ring_id_to_context_switch_event(unsigned int ring_id)
+static int to_context_switch_event(const struct intel_engine_cs *engine)
 {
-	if (WARN_ON(ring_id >= ARRAY_SIZE(context_switch_events)))
+	if (WARN_ON(engine->id >= ARRAY_SIZE(context_switch_events)))
 		return -EINVAL;
 
-	return context_switch_events[ring_id];
+	return context_switch_events[engine->id];
 }
 
 static void switch_virtual_execlist_slot(struct intel_vgpu_execlist *execlist)
@@ -93,9 +92,8 @@ static void emulate_execlist_status(struct intel_vgpu_execlist *execlist)
 	struct execlist_ctx_descriptor_format *desc = execlist->running_context;
 	struct intel_vgpu *vgpu = execlist->vgpu;
 	struct execlist_status_format status;
-	int ring_id = execlist->ring_id;
-	u32 status_reg = execlist_ring_mmio(vgpu->gvt,
-			ring_id, _EL_OFFSET_STATUS);
+	u32 status_reg =
+		execlist_ring_mmio(execlist->engine, _EL_OFFSET_STATUS);
 
 	status.ldw = vgpu_vreg(vgpu, status_reg);
 	status.udw = vgpu_vreg(vgpu, status_reg + 4);
@@ -124,21 +122,19 @@ static void emulate_execlist_status(struct intel_vgpu_execlist *execlist)
 }
 
 static void emulate_csb_update(struct intel_vgpu_execlist *execlist,
-		struct execlist_context_status_format *status,
-		bool trigger_interrupt_later)
+			       struct execlist_context_status_format *status,
+			       bool trigger_interrupt_later)
 {
 	struct intel_vgpu *vgpu = execlist->vgpu;
-	int ring_id = execlist->ring_id;
 	struct execlist_context_status_pointer_format ctx_status_ptr;
 	u32 write_pointer;
 	u32 ctx_status_ptr_reg, ctx_status_buf_reg, offset;
 	unsigned long hwsp_gpa;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 
-	ctx_status_ptr_reg = execlist_ring_mmio(vgpu->gvt, ring_id,
-			_EL_OFFSET_STATUS_PTR);
-	ctx_status_buf_reg = execlist_ring_mmio(vgpu->gvt, ring_id,
-			_EL_OFFSET_STATUS_BUF);
+	ctx_status_ptr_reg =
+		execlist_ring_mmio(execlist->engine, _EL_OFFSET_STATUS_PTR);
+	ctx_status_buf_reg =
+		execlist_ring_mmio(execlist->engine, _EL_OFFSET_STATUS_BUF);
 
 	ctx_status_ptr.dw = vgpu_vreg(vgpu, ctx_status_ptr_reg);
 
@@ -161,26 +157,24 @@ static void emulate_csb_update(struct intel_vgpu_execlist *execlist,
 
 	/* Update the CSB and CSB write pointer in HWSP */
 	hwsp_gpa = intel_vgpu_gma_to_gpa(vgpu->gtt.ggtt_mm,
-					 vgpu->hws_pga[ring_id]);
+					 vgpu->hws_pga[execlist->engine->id]);
 	if (hwsp_gpa != INTEL_GVT_INVALID_ADDR) {
 		intel_gvt_hypervisor_write_gpa(vgpu,
-			hwsp_gpa + I915_HWS_CSB_BUF0_INDEX * 4 +
-			write_pointer * 8,
-			status, 8);
+					       hwsp_gpa + I915_HWS_CSB_BUF0_INDEX * 4 + write_pointer * 8,
+					       status, 8);
 		intel_gvt_hypervisor_write_gpa(vgpu,
-			hwsp_gpa +
-			intel_hws_csb_write_index(dev_priv) * 4,
-			&write_pointer, 4);
+					       hwsp_gpa + intel_hws_csb_write_index(execlist->engine->i915) * 4,
+					       &write_pointer, 4);
 	}
 
 	gvt_dbg_el("vgpu%d: w pointer %u reg %x csb l %x csb h %x\n",
-		vgpu->id, write_pointer, offset, status->ldw, status->udw);
+		   vgpu->id, write_pointer, offset, status->ldw, status->udw);
 
 	if (trigger_interrupt_later)
 		return;
 
 	intel_vgpu_trigger_virtual_event(vgpu,
-			ring_id_to_context_switch_event(execlist->ring_id));
+					 to_context_switch_event(execlist->engine));
 }
 
 static int emulate_execlist_ctx_schedule_out(
@@ -261,9 +255,8 @@ static struct intel_vgpu_execlist_slot *get_next_execlist_slot(
 		struct intel_vgpu_execlist *execlist)
 {
 	struct intel_vgpu *vgpu = execlist->vgpu;
-	int ring_id = execlist->ring_id;
-	u32 status_reg = execlist_ring_mmio(vgpu->gvt, ring_id,
-			_EL_OFFSET_STATUS);
+	u32 status_reg =
+		execlist_ring_mmio(execlist->engine, _EL_OFFSET_STATUS);
 	struct execlist_status_format status;
 
 	status.ldw = vgpu_vreg(vgpu, status_reg);
@@ -379,7 +372,6 @@ static int prepare_execlist_workload(struct intel_vgpu_workload *workload)
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	struct execlist_ctx_descriptor_format ctx[2];
-	int ring_id = workload->ring_id;
 	int ret;
 
 	if (!workload->emulate_schedule_in)
@@ -388,7 +380,8 @@ static int prepare_execlist_workload(struct intel_vgpu_workload *workload)
 	ctx[0] = *get_desc_from_elsp_dwords(&workload->elsp_dwords, 0);
 	ctx[1] = *get_desc_from_elsp_dwords(&workload->elsp_dwords, 1);
 
-	ret = emulate_execlist_schedule_in(&s->execlist[ring_id], ctx);
+	ret = emulate_execlist_schedule_in(&s->execlist[workload->engine->id],
+					   ctx);
 	if (ret) {
 		gvt_vgpu_err("fail to emulate execlist schedule in\n");
 		return ret;
@@ -399,21 +392,21 @@ static int prepare_execlist_workload(struct intel_vgpu_workload *workload)
 static int complete_execlist_workload(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
-	int ring_id = workload->ring_id;
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct intel_vgpu_execlist *execlist = &s->execlist[ring_id];
+	struct intel_vgpu_execlist *execlist =
+		&s->execlist[workload->engine->id];
 	struct intel_vgpu_workload *next_workload;
-	struct list_head *next = workload_q_head(vgpu, ring_id)->next;
+	struct list_head *next = workload_q_head(vgpu, workload->engine)->next;
 	bool lite_restore = false;
 	int ret = 0;
 
-	gvt_dbg_el("complete workload %p status %d\n", workload,
-			workload->status);
+	gvt_dbg_el("complete workload %p status %d\n",
+		   workload, workload->status);
 
-	if (workload->status || (vgpu->resetting_eng & BIT(ring_id)))
+	if (workload->status || vgpu->resetting_eng & workload->engine->mask)
 		goto out;
 
-	if (!list_empty(workload_q_head(vgpu, ring_id))) {
+	if (!list_empty(workload_q_head(vgpu, workload->engine))) {
 		struct execlist_ctx_descriptor_format *this_desc, *next_desc;
 
 		next_workload = container_of(next,
@@ -436,14 +429,15 @@ out:
 	return ret;
 }
 
-static int submit_context(struct intel_vgpu *vgpu, int ring_id,
-		struct execlist_ctx_descriptor_format *desc,
-		bool emulate_schedule_in)
+static int submit_context(struct intel_vgpu *vgpu,
+			  const struct intel_engine_cs *engine,
+			  struct execlist_ctx_descriptor_format *desc,
+			  bool emulate_schedule_in)
 {
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	struct intel_vgpu_workload *workload = NULL;
 
-	workload = intel_vgpu_create_workload(vgpu, ring_id, desc);
+	workload = intel_vgpu_create_workload(vgpu, engine, desc);
 	if (IS_ERR(workload))
 		return PTR_ERR(workload);
 
@@ -452,19 +446,20 @@ static int submit_context(struct intel_vgpu *vgpu, int ring_id,
 	workload->emulate_schedule_in = emulate_schedule_in;
 
 	if (emulate_schedule_in)
-		workload->elsp_dwords = s->execlist[ring_id].elsp_dwords;
+		workload->elsp_dwords = s->execlist[engine->id].elsp_dwords;
 
 	gvt_dbg_el("workload %p emulate schedule_in %d\n", workload,
-			emulate_schedule_in);
+		   emulate_schedule_in);
 
 	intel_vgpu_queue_workload(workload);
 	return 0;
 }
 
-int intel_vgpu_submit_execlist(struct intel_vgpu *vgpu, int ring_id)
+int intel_vgpu_submit_execlist(struct intel_vgpu *vgpu,
+			       const struct intel_engine_cs *engine)
 {
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct intel_vgpu_execlist *execlist = &s->execlist[ring_id];
+	struct intel_vgpu_execlist *execlist = &s->execlist[engine->id];
 	struct execlist_ctx_descriptor_format *desc[2];
 	int i, ret;
 
@@ -489,7 +484,7 @@ int intel_vgpu_submit_execlist(struct intel_vgpu *vgpu, int ring_id)
 	for (i = 0; i < ARRAY_SIZE(desc); i++) {
 		if (!desc[i]->valid)
 			continue;
-		ret = submit_context(vgpu, ring_id, desc[i], i == 0);
+		ret = submit_context(vgpu, engine, desc[i], i == 0);
 		if (ret) {
 			gvt_vgpu_err("failed to submit desc %d\n", i);
 			return ret;
@@ -504,22 +499,22 @@ inv_desc:
 	return -EINVAL;
 }
 
-static void init_vgpu_execlist(struct intel_vgpu *vgpu, int ring_id)
+static void init_vgpu_execlist(struct intel_vgpu *vgpu,
+			       const struct intel_engine_cs *engine)
 {
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct intel_vgpu_execlist *execlist = &s->execlist[ring_id];
+	struct intel_vgpu_execlist *execlist = &s->execlist[engine->id];
 	struct execlist_context_status_pointer_format ctx_status_ptr;
 	u32 ctx_status_ptr_reg;
 
 	memset(execlist, 0, sizeof(*execlist));
 
 	execlist->vgpu = vgpu;
-	execlist->ring_id = ring_id;
+	execlist->engine = engine;
 	execlist->slot[0].index = 0;
 	execlist->slot[1].index = 1;
 
-	ctx_status_ptr_reg = execlist_ring_mmio(vgpu->gvt, ring_id,
-			_EL_OFFSET_STATUS_PTR);
+	ctx_status_ptr_reg = execlist_ring_mmio(engine, _EL_OFFSET_STATUS_PTR);
 	ctx_status_ptr.dw = vgpu_vreg(vgpu, ctx_status_ptr_reg);
 	ctx_status_ptr.read_ptr = 0;
 	ctx_status_ptr.write_ptr = 0x7;
@@ -529,7 +524,7 @@ static void init_vgpu_execlist(struct intel_vgpu *vgpu, int ring_id)
 static void clean_execlist(struct intel_vgpu *vgpu,
 			   intel_engine_mask_t engine_mask)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	struct intel_engine_cs *engine;
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	intel_engine_mask_t tmp;
@@ -544,12 +539,12 @@ static void clean_execlist(struct intel_vgpu *vgpu,
 static void reset_execlist(struct intel_vgpu *vgpu,
 			   intel_engine_mask_t engine_mask)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	struct intel_engine_cs *engine;
 	intel_engine_mask_t tmp;
 
 	for_each_engine_masked(engine, &dev_priv->gt, engine_mask, tmp)
-		init_vgpu_execlist(vgpu, engine->id);
+		init_vgpu_execlist(vgpu, engine);
 }
 
 static int init_execlist(struct intel_vgpu *vgpu,
diff --git a/drivers/gpu/drm/i915/gvt/execlist.h b/drivers/gpu/drm/i915/gvt/execlist.h
index 5c0c1fd30c83..d62cd14605a3 100644
--- a/drivers/gpu/drm/i915/gvt/execlist.h
+++ b/drivers/gpu/drm/i915/gvt/execlist.h
@@ -170,16 +170,17 @@ struct intel_vgpu_execlist {
 	struct intel_vgpu_execlist_slot *running_slot;
 	struct intel_vgpu_execlist_slot *pending_slot;
 	struct execlist_ctx_descriptor_format *running_context;
-	int ring_id;
 	struct intel_vgpu *vgpu;
 	struct intel_vgpu_elsp_dwords elsp_dwords;
+	const struct intel_engine_cs *engine;
 };
 
 void intel_vgpu_clean_execlist(struct intel_vgpu *vgpu);
 
 int intel_vgpu_init_execlist(struct intel_vgpu *vgpu);
 
-int intel_vgpu_submit_execlist(struct intel_vgpu *vgpu, int ring_id);
+int intel_vgpu_submit_execlist(struct intel_vgpu *vgpu,
+			       const struct intel_engine_cs *engine);
 
 void intel_vgpu_reset_execlist(struct intel_vgpu *vgpu,
 			       intel_engine_mask_t engine_mask);
diff --git a/drivers/gpu/drm/i915/gvt/fb_decoder.c b/drivers/gpu/drm/i915/gvt/fb_decoder.c
index 8bb292b01271..0889ad8291b0 100644
--- a/drivers/gpu/drm/i915/gvt/fb_decoder.c
+++ b/drivers/gpu/drm/i915/gvt/fb_decoder.c
@@ -146,7 +146,7 @@ static int skl_format_to_drm(int format, bool rgb_order, bool alpha,
 static u32 intel_vgpu_get_stride(struct intel_vgpu *vgpu, int pipe,
 	u32 tiled, int stride_mask, int bpp)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 
 	u32 stride_reg = vgpu_vreg_t(vgpu, DSPSTRIDE(pipe)) & stride_mask;
 	u32 stride = stride_reg;
@@ -202,8 +202,8 @@ static int get_active_pipe(struct intel_vgpu *vgpu)
 int intel_vgpu_decode_primary_plane(struct intel_vgpu *vgpu,
 	struct intel_vgpu_primary_plane_format *plane)
 {
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	u32 val, fmt;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	int pipe;
 
 	pipe = get_active_pipe(vgpu);
@@ -332,9 +332,9 @@ static int cursor_mode_to_drm(int mode)
 int intel_vgpu_decode_cursor_plane(struct intel_vgpu *vgpu,
 	struct intel_vgpu_cursor_plane_format *plane)
 {
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	u32 val, mode, index;
 	u32 alpha_plane, alpha_force;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	int pipe;
 
 	pipe = get_active_pipe(vgpu);
diff --git a/drivers/gpu/drm/i915/gvt/firmware.c b/drivers/gpu/drm/i915/gvt/firmware.c
index b0c1fda32977..990a181094e3 100644
--- a/drivers/gpu/drm/i915/gvt/firmware.c
+++ b/drivers/gpu/drm/i915/gvt/firmware.c
@@ -68,9 +68,7 @@ static struct bin_attribute firmware_attr = {
 
 static int mmio_snapshot_handler(struct intel_gvt *gvt, u32 offset, void *data)
 {
-	struct drm_i915_private *i915 = gvt->dev_priv;
-
-	*(u32 *)(data + offset) = intel_uncore_read_notrace(&i915->uncore,
+	*(u32 *)(data + offset) = intel_uncore_read_notrace(gvt->gt->uncore,
 							    _MMIO(offset));
 	return 0;
 }
@@ -78,7 +76,7 @@ static int mmio_snapshot_handler(struct intel_gvt *gvt, u32 offset, void *data)
 static int expose_firmware_sysfs(struct intel_gvt *gvt)
 {
 	struct intel_gvt_device_info *info = &gvt->device_info;
-	struct pci_dev *pdev = gvt->dev_priv->drm.pdev;
+	struct pci_dev *pdev = gvt->gt->i915->drm.pdev;
 	struct gvt_firmware_header *h;
 	void *firmware;
 	void *p;
@@ -129,7 +127,7 @@ static int expose_firmware_sysfs(struct intel_gvt *gvt)
 
 static void clean_firmware_sysfs(struct intel_gvt *gvt)
 {
-	struct pci_dev *pdev = gvt->dev_priv->drm.pdev;
+	struct pci_dev *pdev = gvt->gt->i915->drm.pdev;
 
 	device_remove_bin_file(&pdev->dev, &firmware_attr);
 	vfree(firmware_attr.private);
@@ -153,8 +151,7 @@ static int verify_firmware(struct intel_gvt *gvt,
 			   const struct firmware *fw)
 {
 	struct intel_gvt_device_info *info = &gvt->device_info;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
-	struct pci_dev *pdev = dev_priv->drm.pdev;
+	struct pci_dev *pdev = gvt->gt->i915->drm.pdev;
 	struct gvt_firmware_header *h;
 	unsigned long id, crc32_start;
 	const void *mem;
@@ -208,8 +205,7 @@ invalid_firmware:
 int intel_gvt_load_firmware(struct intel_gvt *gvt)
 {
 	struct intel_gvt_device_info *info = &gvt->device_info;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
-	struct pci_dev *pdev = dev_priv->drm.pdev;
+	struct pci_dev *pdev = gvt->gt->i915->drm.pdev;
 	struct intel_gvt_firmware *firmware = &gvt->firmware;
 	struct gvt_firmware_header *h;
 	const struct firmware *fw;
@@ -244,7 +240,7 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt)
 
 	gvt_dbg_core("request hw state firmware %s...\n", path);
 
-	ret = request_firmware(&fw, path, &dev_priv->drm.pdev->dev);
+	ret = request_firmware(&fw, path, &gvt->gt->i915->drm.pdev->dev);
 	kfree(path);
 
 	if (ret)
diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
index 4a4828074cb7..2a4b23f8aa74 100644
--- a/drivers/gpu/drm/i915/gvt/gtt.c
+++ b/drivers/gpu/drm/i915/gvt/gtt.c
@@ -71,8 +71,10 @@ bool intel_gvt_ggtt_validate_range(struct intel_vgpu *vgpu, u64 addr, u32 size)
 /* translate a guest gmadr to host gmadr */
 int intel_gvt_ggtt_gmadr_g2h(struct intel_vgpu *vgpu, u64 g_addr, u64 *h_addr)
 {
-	if (WARN(!vgpu_gmadr_is_valid(vgpu, g_addr),
-		 "invalid guest gmadr %llx\n", g_addr))
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+
+	if (drm_WARN(&i915->drm, !vgpu_gmadr_is_valid(vgpu, g_addr),
+		     "invalid guest gmadr %llx\n", g_addr))
 		return -EACCES;
 
 	if (vgpu_gmadr_is_aperture(vgpu, g_addr))
@@ -87,8 +89,10 @@ int intel_gvt_ggtt_gmadr_g2h(struct intel_vgpu *vgpu, u64 g_addr, u64 *h_addr)
 /* translate a host gmadr to guest gmadr */
 int intel_gvt_ggtt_gmadr_h2g(struct intel_vgpu *vgpu, u64 h_addr, u64 *g_addr)
 {
-	if (WARN(!gvt_gmadr_is_valid(vgpu->gvt, h_addr),
-		 "invalid host gmadr %llx\n", h_addr))
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+
+	if (drm_WARN(&i915->drm, !gvt_gmadr_is_valid(vgpu->gvt, h_addr),
+		     "invalid host gmadr %llx\n", h_addr))
 		return -EACCES;
 
 	if (gvt_gmadr_is_aperture(vgpu->gvt, h_addr))
@@ -275,24 +279,23 @@ static inline int get_pse_type(int type)
 	return gtt_type_table[type].pse_entry_type;
 }
 
-static u64 read_pte64(struct drm_i915_private *dev_priv, unsigned long index)
+static u64 read_pte64(struct i915_ggtt *ggtt, unsigned long index)
 {
-	void __iomem *addr = (gen8_pte_t __iomem *)dev_priv->ggtt.gsm + index;
+	void __iomem *addr = (gen8_pte_t __iomem *)ggtt->gsm + index;
 
 	return readq(addr);
 }
 
-static void ggtt_invalidate(struct drm_i915_private *dev_priv)
+static void ggtt_invalidate(struct intel_gt *gt)
 {
-	mmio_hw_access_pre(dev_priv);
-	I915_WRITE(GFX_FLSH_CNTL_GEN6, GFX_FLSH_CNTL_EN);
-	mmio_hw_access_post(dev_priv);
+	mmio_hw_access_pre(gt);
+	intel_uncore_write(gt->uncore, GFX_FLSH_CNTL_GEN6, GFX_FLSH_CNTL_EN);
+	mmio_hw_access_post(gt);
 }
 
-static void write_pte64(struct drm_i915_private *dev_priv,
-		unsigned long index, u64 pte)
+static void write_pte64(struct i915_ggtt *ggtt, unsigned long index, u64 pte)
 {
-	void __iomem *addr = (gen8_pte_t __iomem *)dev_priv->ggtt.gsm + index;
+	void __iomem *addr = (gen8_pte_t __iomem *)ggtt->gsm + index;
 
 	writeq(pte, addr);
 }
@@ -315,7 +318,7 @@ static inline int gtt_get_entry64(void *pt,
 		if (WARN_ON(ret))
 			return ret;
 	} else if (!pt) {
-		e->val64 = read_pte64(vgpu->gvt->dev_priv, index);
+		e->val64 = read_pte64(vgpu->gvt->gt->ggtt, index);
 	} else {
 		e->val64 = *((u64 *)pt + index);
 	}
@@ -340,7 +343,7 @@ static inline int gtt_set_entry64(void *pt,
 		if (WARN_ON(ret))
 			return ret;
 	} else if (!pt) {
-		write_pte64(vgpu->gvt->dev_priv, index, e->val64);
+		write_pte64(vgpu->gvt->gt->ggtt, index, e->val64);
 	} else {
 		*((u64 *)pt + index) = e->val64;
 	}
@@ -734,7 +737,7 @@ static int detach_oos_page(struct intel_vgpu *vgpu,
 
 static void ppgtt_free_spt(struct intel_vgpu_ppgtt_spt *spt)
 {
-	struct device *kdev = &spt->vgpu->gvt->dev_priv->drm.pdev->dev;
+	struct device *kdev = &spt->vgpu->gvt->gt->i915->drm.pdev->dev;
 
 	trace_spt_free(spt->vgpu->id, spt, spt->guest_page.type);
 
@@ -819,7 +822,7 @@ static int reclaim_one_ppgtt_mm(struct intel_gvt *gvt);
 static struct intel_vgpu_ppgtt_spt *ppgtt_alloc_spt(
 		struct intel_vgpu *vgpu, enum intel_gvt_gtt_type type)
 {
-	struct device *kdev = &vgpu->gvt->dev_priv->drm.pdev->dev;
+	struct device *kdev = &vgpu->gvt->gt->i915->drm.pdev->dev;
 	struct intel_vgpu_ppgtt_spt *spt = NULL;
 	dma_addr_t daddr;
 	int ret;
@@ -940,6 +943,7 @@ static int ppgtt_invalidate_spt(struct intel_vgpu_ppgtt_spt *spt);
 static int ppgtt_invalidate_spt_by_shadow_entry(struct intel_vgpu *vgpu,
 		struct intel_gvt_gtt_entry *e)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_gvt_gtt_pte_ops *ops = vgpu->gvt->gtt.pte_ops;
 	struct intel_vgpu_ppgtt_spt *s;
 	enum intel_gvt_gtt_type cur_pt_type;
@@ -952,7 +956,9 @@ static int ppgtt_invalidate_spt_by_shadow_entry(struct intel_vgpu *vgpu,
 
 		if (!gtt_type_is_pt(cur_pt_type) ||
 				!gtt_type_is_pt(cur_pt_type + 1)) {
-			WARN(1, "Invalid page table type, cur_pt_type is: %d\n", cur_pt_type);
+			drm_WARN(&i915->drm, 1,
+				 "Invalid page table type, cur_pt_type is: %d\n",
+				 cur_pt_type);
 			return -EINVAL;
 		}
 
@@ -1044,7 +1050,7 @@ fail:
 
 static bool vgpu_ips_enabled(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 
 	if (INTEL_GEN(dev_priv) == 9 || INTEL_GEN(dev_priv) == 10) {
 		u32 ips = vgpu_vreg_t(vgpu, GEN8_GAMW_ECO_DEV_RW_IA) &
@@ -1153,7 +1159,7 @@ static int is_2MB_gtt_possible(struct intel_vgpu *vgpu,
 	struct intel_gvt_gtt_pte_ops *ops = vgpu->gvt->gtt.pte_ops;
 	unsigned long pfn;
 
-	if (!HAS_PAGE_SIZES(vgpu->gvt->dev_priv, I915_GTT_PAGE_SIZE_2M))
+	if (!HAS_PAGE_SIZES(vgpu->gvt->gt->i915, I915_GTT_PAGE_SIZE_2M))
 		return 0;
 
 	pfn = intel_gvt_hypervisor_gfn_to_mfn(vgpu, ops->get_pfn(entry));
@@ -2314,7 +2320,7 @@ out:
 	ggtt_invalidate_pte(vgpu, &e);
 
 	ggtt_set_host_entry(ggtt_mm, &m, g_gtt_index);
-	ggtt_invalidate(gvt->dev_priv);
+	ggtt_invalidate(gvt->gt);
 	return 0;
 }
 
@@ -2347,16 +2353,18 @@ int intel_vgpu_emulate_ggtt_mmio_write(struct intel_vgpu *vgpu,
 static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 		enum intel_gvt_gtt_type type)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_vgpu_gtt *gtt = &vgpu->gtt;
 	struct intel_gvt_gtt_pte_ops *ops = vgpu->gvt->gtt.pte_ops;
 	int page_entry_num = I915_GTT_PAGE_SIZE >>
 				vgpu->gvt->device_info.gtt_entry_size_shift;
 	void *scratch_pt;
 	int i;
-	struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev;
+	struct device *dev = &vgpu->gvt->gt->i915->drm.pdev->dev;
 	dma_addr_t daddr;
 
-	if (WARN_ON(type < GTT_TYPE_PPGTT_PTE_PT || type >= GTT_TYPE_MAX))
+	if (drm_WARN_ON(&i915->drm,
+			type < GTT_TYPE_PPGTT_PTE_PT || type >= GTT_TYPE_MAX))
 		return -EINVAL;
 
 	scratch_pt = (void *)get_zeroed_page(GFP_KERNEL);
@@ -2410,7 +2418,7 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
 static int release_scratch_page_tree(struct intel_vgpu *vgpu)
 {
 	int i;
-	struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev;
+	struct device *dev = &vgpu->gvt->gt->i915->drm.pdev->dev;
 	dma_addr_t daddr;
 
 	for (i = GTT_TYPE_PPGTT_PTE_PT; i < GTT_TYPE_MAX; i++) {
@@ -2682,7 +2690,7 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt)
 {
 	int ret;
 	void *page;
-	struct device *dev = &gvt->dev_priv->drm.pdev->dev;
+	struct device *dev = &gvt->gt->i915->drm.pdev->dev;
 	dma_addr_t daddr;
 
 	gvt_dbg_core("init gtt\n");
@@ -2731,7 +2739,7 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt)
  */
 void intel_gvt_clean_gtt(struct intel_gvt *gvt)
 {
-	struct device *dev = &gvt->dev_priv->drm.pdev->dev;
+	struct device *dev = &gvt->gt->i915->drm.pdev->dev;
 	dma_addr_t daddr = (dma_addr_t)(gvt->gtt.scratch_mfn <<
 					I915_GTT_PAGE_SHIFT);
 
@@ -2779,7 +2787,6 @@ void intel_vgpu_invalidate_ppgtt(struct intel_vgpu *vgpu)
 void intel_vgpu_reset_ggtt(struct intel_vgpu *vgpu, bool invalidate_old)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
 	struct intel_gvt_gtt_pte_ops *pte_ops = vgpu->gvt->gtt.pte_ops;
 	struct intel_gvt_gtt_entry entry = {.type = GTT_TYPE_GGTT_PTE};
 	struct intel_gvt_gtt_entry old_entry;
@@ -2809,7 +2816,7 @@ void intel_vgpu_reset_ggtt(struct intel_vgpu *vgpu, bool invalidate_old)
 		ggtt_set_host_entry(vgpu->gtt.ggtt_mm, &entry, index++);
 	}
 
-	ggtt_invalidate(dev_priv);
+	ggtt_invalidate(gvt->gt);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/gvt/gvt.c b/drivers/gpu/drm/i915/gvt/gvt.c
index 8f37eefa0a02..9e1787867894 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.c
+++ b/drivers/gpu/drm/i915/gvt/gvt.c
@@ -35,6 +35,7 @@
 #include <linux/kthread.h>
 
 #include "i915_drv.h"
+#include "intel_gvt.h"
 #include "gvt.h"
 #include <linux/vfio.h>
 #include <linux/mdev.h>
@@ -49,15 +50,15 @@ static const char * const supported_hypervisors[] = {
 static struct intel_vgpu_type *intel_gvt_find_vgpu_type(struct intel_gvt *gvt,
 		const char *name)
 {
+	const char *driver_name =
+		dev_driver_string(&gvt->gt->i915->drm.pdev->dev);
 	int i;
-	struct intel_vgpu_type *t;
-	const char *driver_name = dev_driver_string(
-			&gvt->dev_priv->drm.pdev->dev);
 
+	name += strlen(driver_name) + 1;
 	for (i = 0; i < gvt->num_types; i++) {
-		t = &gvt->types[i];
-		if (!strncmp(t->name, name + strlen(driver_name) + 1,
-			sizeof(t->name)))
+		struct intel_vgpu_type *t = &gvt->types[i];
+
+		if (!strncmp(t->name, name, sizeof(t->name)))
 			return t;
 	}
 
@@ -120,10 +121,8 @@ static struct attribute_group *gvt_vgpu_type_groups[] = {
 	[0 ... NR_MAX_INTEL_VGPU_TYPES - 1] = NULL,
 };
 
-static bool intel_get_gvt_attrs(struct attribute ***type_attrs,
-		struct attribute_group ***intel_vgpu_type_groups)
+static bool intel_get_gvt_attrs(struct attribute_group ***intel_vgpu_type_groups)
 {
-	*type_attrs = gvt_type_attrs;
 	*intel_vgpu_type_groups = gvt_vgpu_type_groups;
 	return true;
 }
@@ -191,7 +190,7 @@ static const struct intel_gvt_ops intel_gvt_ops = {
 static void init_device_info(struct intel_gvt *gvt)
 {
 	struct intel_gvt_device_info *info = &gvt->device_info;
-	struct pci_dev *pdev = gvt->dev_priv->drm.pdev;
+	struct pci_dev *pdev = gvt->gt->i915->drm.pdev;
 
 	info->max_support_vgpus = 8;
 	info->cfg_space_size = PCI_CFG_SPACE_EXP_SIZE;
@@ -257,17 +256,17 @@ static int init_service_thread(struct intel_gvt *gvt)
 
 /**
  * intel_gvt_clean_device - clean a GVT device
- * @dev_priv: i915 private
+ * @i915: i915 private
  *
  * This function is called at the driver unloading stage, to free the
  * resources owned by a GVT device.
  *
  */
-void intel_gvt_clean_device(struct drm_i915_private *dev_priv)
+void intel_gvt_clean_device(struct drm_i915_private *i915)
 {
-	struct intel_gvt *gvt = to_gvt(dev_priv);
+	struct intel_gvt *gvt = fetch_and_zero(&i915->gvt);
 
-	if (WARN_ON(!gvt))
+	if (drm_WARN_ON(&i915->drm, !gvt))
 		return;
 
 	intel_gvt_destroy_idle_vgpu(gvt->idle_vgpu);
@@ -285,13 +284,12 @@ void intel_gvt_clean_device(struct drm_i915_private *dev_priv)
 	intel_gvt_clean_mmio_info(gvt);
 	idr_destroy(&gvt->vgpu_idr);
 
-	kfree(dev_priv->gvt);
-	dev_priv->gvt = NULL;
+	kfree(i915->gvt);
 }
 
 /**
  * intel_gvt_init_device - initialize a GVT device
- * @dev_priv: drm i915 private data
+ * @i915: drm i915 private data
  *
  * This function is called at the initialization stage, to initialize
  * necessary GVT components.
@@ -300,13 +298,13 @@ void intel_gvt_clean_device(struct drm_i915_private *dev_priv)
  * Zero on success, negative error code if failed.
  *
  */
-int intel_gvt_init_device(struct drm_i915_private *dev_priv)
+int intel_gvt_init_device(struct drm_i915_private *i915)
 {
 	struct intel_gvt *gvt;
 	struct intel_vgpu *vgpu;
 	int ret;
 
-	if (WARN_ON(dev_priv->gvt))
+	if (drm_WARN_ON(&i915->drm, i915->gvt))
 		return -EEXIST;
 
 	gvt = kzalloc(sizeof(struct intel_gvt), GFP_KERNEL);
@@ -319,7 +317,8 @@ int intel_gvt_init_device(struct drm_i915_private *dev_priv)
 	spin_lock_init(&gvt->scheduler.mmio_context_lock);
 	mutex_init(&gvt->lock);
 	mutex_init(&gvt->sched_lock);
-	gvt->dev_priv = dev_priv;
+	gvt->gt = &i915->gt;
+	i915->gvt = gvt;
 
 	init_device_info(gvt);
 
@@ -378,8 +377,7 @@ int intel_gvt_init_device(struct drm_i915_private *dev_priv)
 	intel_gvt_debugfs_init(gvt);
 
 	gvt_dbg_core("gvt device initialization is done\n");
-	dev_priv->gvt = gvt;
-	intel_gvt_host.dev = &dev_priv->drm.pdev->dev;
+	intel_gvt_host.dev = &i915->drm.pdev->dev;
 	intel_gvt_host.initialized = true;
 	return 0;
 
@@ -404,6 +402,7 @@ out_clean_mmio_info:
 out_clean_idr:
 	idr_destroy(&gvt->vgpu_idr);
 	kfree(gvt);
+	i915->gvt = NULL;
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h
index b47c6acaf9c0..58c2c7932e3f 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.h
+++ b/drivers/gpu/drm/i915/gvt/gvt.h
@@ -196,41 +196,21 @@ struct intel_vgpu {
 
 	struct dentry *debugfs;
 
-#if IS_ENABLED(CONFIG_DRM_I915_GVT_KVMGT)
-	struct {
-		struct mdev_device *mdev;
-		struct vfio_region *region;
-		int num_regions;
-		struct eventfd_ctx *intx_trigger;
-		struct eventfd_ctx *msi_trigger;
-
-		/*
-		 * Two caches are used to avoid mapping duplicated pages (eg.
-		 * scratch pages). This help to reduce dma setup overhead.
-		 */
-		struct rb_root gfn_cache;
-		struct rb_root dma_addr_cache;
-		unsigned long nr_cache_entries;
-		struct mutex cache_lock;
-
-		struct notifier_block iommu_notifier;
-		struct notifier_block group_notifier;
-		struct kvm *kvm;
-		struct work_struct release_work;
-		atomic_t released;
-		struct vfio_device *vfio_device;
-	} vdev;
-#endif
+	/* Hypervisor-specific device state. */
+	void *vdev;
 
 	struct list_head dmabuf_obj_list_head;
 	struct mutex dmabuf_lock;
 	struct idr object_idr;
 
-	struct completion vblank_done;
-
 	u32 scan_nonprivbb;
 };
 
+static inline void *intel_vgpu_vdev(struct intel_vgpu *vgpu)
+{
+	return vgpu->vdev;
+}
+
 /* validating GM healthy status*/
 #define vgpu_is_vm_unhealthy(ret_val) \
 	(((ret_val) == -EBADRQC) || ((ret_val) == -EFAULT))
@@ -306,7 +286,7 @@ struct intel_gvt {
 	/* scheduler scope lock, protect gvt and vgpu schedule related data */
 	struct mutex sched_lock;
 
-	struct drm_i915_private *dev_priv;
+	struct intel_gt *gt;
 	struct idr vgpu_idr;	/* vGPU IDR pool */
 
 	struct intel_gvt_device_info device_info;
@@ -376,14 +356,15 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt);
 #define HOST_HIGH_GM_SIZE MB_TO_BYTES(384)
 #define HOST_FENCE 4
 
+#define gvt_to_ggtt(gvt)	((gvt)->gt->ggtt)
+
 /* Aperture/GM space definitions for GVT device */
-#define gvt_aperture_sz(gvt)	  (gvt->dev_priv->ggtt.mappable_end)
-#define gvt_aperture_pa_base(gvt) (gvt->dev_priv->ggtt.gmadr.start)
+#define gvt_aperture_sz(gvt)	  gvt_to_ggtt(gvt)->mappable_end
+#define gvt_aperture_pa_base(gvt) gvt_to_ggtt(gvt)->gmadr.start
 
-#define gvt_ggtt_gm_sz(gvt)	  (gvt->dev_priv->ggtt.vm.total)
-#define gvt_ggtt_sz(gvt) \
-	((gvt->dev_priv->ggtt.vm.total >> PAGE_SHIFT) << 3)
-#define gvt_hidden_sz(gvt)	  (gvt_ggtt_gm_sz(gvt) - gvt_aperture_sz(gvt))
+#define gvt_ggtt_gm_sz(gvt)	gvt_to_ggtt(gvt)->vm.total
+#define gvt_ggtt_sz(gvt)	(gvt_to_ggtt(gvt)->vm.total >> PAGE_SHIFT << 3)
+#define gvt_hidden_sz(gvt)	(gvt_ggtt_gm_sz(gvt) - gvt_aperture_sz(gvt))
 
 #define gvt_aperture_gmadr_base(gvt) (0)
 #define gvt_aperture_gmadr_end(gvt) (gvt_aperture_gmadr_base(gvt) \
@@ -394,7 +375,7 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt);
 #define gvt_hidden_gmadr_end(gvt) (gvt_hidden_gmadr_base(gvt) \
 				   + gvt_hidden_sz(gvt) - 1)
 
-#define gvt_fence_sz(gvt) ((gvt)->dev_priv->ggtt.num_fences)
+#define gvt_fence_sz(gvt) (gvt_to_ggtt(gvt)->num_fences)
 
 /* Aperture/GM space definitions for vGPU */
 #define vgpu_aperture_offset(vgpu)	((vgpu)->gm.low_gm_node.start)
@@ -570,8 +551,7 @@ struct intel_gvt_ops {
 	void (*vgpu_deactivate)(struct intel_vgpu *);
 	struct intel_vgpu_type *(*gvt_find_vgpu_type)(struct intel_gvt *gvt,
 			const char *name);
-	bool (*get_gvt_attrs)(struct attribute ***type_attrs,
-			struct attribute_group ***intel_vgpu_type_groups);
+	bool (*get_gvt_attrs)(struct attribute_group ***intel_vgpu_type_groups);
 	int (*vgpu_query_plane)(struct intel_vgpu *vgpu, void *);
 	int (*vgpu_get_dmabuf)(struct intel_vgpu *vgpu, unsigned int);
 	int (*write_protect_handler)(struct intel_vgpu *, u64, void *,
@@ -586,14 +566,14 @@ enum {
 	GVT_FAILSAFE_GUEST_ERR,
 };
 
-static inline void mmio_hw_access_pre(struct drm_i915_private *dev_priv)
+static inline void mmio_hw_access_pre(struct intel_gt *gt)
 {
-	intel_runtime_pm_get(&dev_priv->runtime_pm);
+	intel_runtime_pm_get(gt->uncore->rpm);
 }
 
-static inline void mmio_hw_access_post(struct drm_i915_private *dev_priv)
+static inline void mmio_hw_access_post(struct intel_gt *gt)
 {
-	intel_runtime_pm_put_unchecked(&dev_priv->runtime_pm);
+	intel_runtime_pm_put_unchecked(gt->uncore->rpm);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index ae6700dc9d73..0182e2a5acff 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -49,15 +49,17 @@
 
 unsigned long intel_gvt_get_device_type(struct intel_gvt *gvt)
 {
-	if (IS_BROADWELL(gvt->dev_priv))
+	struct drm_i915_private *i915 = gvt->gt->i915;
+
+	if (IS_BROADWELL(i915))
 		return D_BDW;
-	else if (IS_SKYLAKE(gvt->dev_priv))
+	else if (IS_SKYLAKE(i915))
 		return D_SKL;
-	else if (IS_KABYLAKE(gvt->dev_priv))
+	else if (IS_KABYLAKE(i915))
 		return D_KBL;
-	else if (IS_BROXTON(gvt->dev_priv))
+	else if (IS_BROXTON(i915))
 		return D_BXT;
-	else if (IS_COFFEELAKE(gvt->dev_priv))
+	else if (IS_COFFEELAKE(i915))
 		return D_CFL;
 
 	return 0;
@@ -142,25 +144,25 @@ static int new_mmio_info(struct intel_gvt *gvt,
 }
 
 /**
- * intel_gvt_render_mmio_to_ring_id - convert a mmio offset into ring id
+ * intel_gvt_render_mmio_to_engine - convert a mmio offset into the engine
  * @gvt: a GVT device
  * @offset: register offset
  *
  * Returns:
- * Ring ID on success, negative error code if failed.
+ * The engine containing the offset within its mmio page.
  */
-int intel_gvt_render_mmio_to_ring_id(struct intel_gvt *gvt,
-		unsigned int offset)
+const struct intel_engine_cs *
+intel_gvt_render_mmio_to_engine(struct intel_gvt *gvt, unsigned int offset)
 {
-	enum intel_engine_id id;
 	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
 
 	offset &= ~GENMASK(11, 0);
-	for_each_engine(engine, gvt->dev_priv, id) {
+	for_each_engine(engine, gvt->gt, id)
 		if (engine->mmio_base == offset)
-			return id;
-	}
-	return -ENODEV;
+			return engine;
+
+	return NULL;
 }
 
 #define offset_to_fence_num(offset) \
@@ -217,7 +219,7 @@ static int gamw_echo_dev_rw_ia_write(struct intel_vgpu *vgpu,
 {
 	u32 ips = (*(u32 *)p_data) & GAMW_ECO_ENABLE_64K_IPS_FIELD;
 
-	if (INTEL_GEN(vgpu->gvt->dev_priv) <= 10) {
+	if (INTEL_GEN(vgpu->gvt->gt->i915) <= 10) {
 		if (ips == GAMW_ECO_ENABLE_64K_IPS_FIELD)
 			gvt_dbg_core("vgpu%d: ips enabled\n", vgpu->id);
 		else if (!ips)
@@ -253,7 +255,7 @@ static int fence_mmio_read(struct intel_vgpu *vgpu, unsigned int off,
 static int fence_mmio_write(struct intel_vgpu *vgpu, unsigned int off,
 		void *p_data, unsigned int bytes)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct intel_gvt *gvt = vgpu->gvt;
 	unsigned int fence_num = offset_to_fence_num(off);
 	int ret;
 
@@ -262,10 +264,10 @@ static int fence_mmio_write(struct intel_vgpu *vgpu, unsigned int off,
 		return ret;
 	write_vreg(vgpu, off, p_data, bytes);
 
-	mmio_hw_access_pre(dev_priv);
+	mmio_hw_access_pre(gvt->gt);
 	intel_vgpu_write_fence(vgpu, fence_num,
 			vgpu_vreg64(vgpu, fence_num_to_offset(fence_num)));
-	mmio_hw_access_post(dev_priv);
+	mmio_hw_access_post(gvt->gt);
 	return 0;
 }
 
@@ -283,7 +285,7 @@ static int mul_force_wake_write(struct intel_vgpu *vgpu,
 	old = vgpu_vreg(vgpu, offset);
 	new = CALC_MODE_MASK_REG(old, *(u32 *)p_data);
 
-	if (INTEL_GEN(vgpu->gvt->dev_priv)  >=  9) {
+	if (INTEL_GEN(vgpu->gvt->gt->i915)  >=  9) {
 		switch (offset) {
 		case FORCEWAKE_RENDER_GEN9_REG:
 			ack_reg_offset = FORCEWAKE_ACK_RENDER_GEN9_REG;
@@ -345,7 +347,7 @@ static int gdrst_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 			gvt_dbg_mmio("vgpu%d: request GUC Reset\n", vgpu->id);
 			vgpu_vreg_t(vgpu, GUC_STATUS) |= GS_MIA_IN_RESET;
 		}
-		engine_mask &= INTEL_INFO(vgpu->gvt->dev_priv)->engine_mask;
+		engine_mask &= INTEL_INFO(vgpu->gvt->gt->i915)->engine_mask;
 	}
 
 	/* vgpu_lock already hold by emulate mmio r/w */
@@ -492,7 +494,7 @@ static i915_reg_t force_nonpriv_white_list[] = {
 };
 
 /* a simple bsearch */
-static inline bool in_whitelist(unsigned int reg)
+static inline bool in_whitelist(u32 reg)
 {
 	int left = 0, right = ARRAY_SIZE(force_nonpriv_white_list);
 	i915_reg_t *array = force_nonpriv_white_list;
@@ -514,26 +516,21 @@ static int force_nonpriv_write(struct intel_vgpu *vgpu,
 	unsigned int offset, void *p_data, unsigned int bytes)
 {
 	u32 reg_nonpriv = (*(u32 *)p_data) & REG_GENMASK(25, 2);
-	int ring_id = intel_gvt_render_mmio_to_ring_id(vgpu->gvt, offset);
-	u32 ring_base;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
-	int ret = -EINVAL;
-
-	if ((bytes != 4) || ((offset & (bytes - 1)) != 0) || ring_id < 0) {
-		gvt_err("vgpu(%d) ring %d Invalid FORCE_NONPRIV offset %x(%dB)\n",
-			vgpu->id, ring_id, offset, bytes);
-		return ret;
-	}
+	const struct intel_engine_cs *engine =
+		intel_gvt_render_mmio_to_engine(vgpu->gvt, offset);
 
-	ring_base = dev_priv->engine[ring_id]->mmio_base;
+	if (bytes != 4 || !IS_ALIGNED(offset, bytes) || !engine) {
+		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV offset %x(%dB)\n",
+			vgpu->id, offset, bytes);
+		return -EINVAL;
+	}
 
-	if (in_whitelist(reg_nonpriv) ||
-		reg_nonpriv == i915_mmio_reg_offset(RING_NOPID(ring_base))) {
-		ret = intel_vgpu_default_mmio_write(vgpu, offset, p_data,
-			bytes);
-	} else
+	if (!in_whitelist(reg_nonpriv) &&
+	    reg_nonpriv != i915_mmio_reg_offset(RING_NOPID(engine->mmio_base))) {
 		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x at offset %x\n",
-			vgpu->id, *(u32 *)p_data, offset);
+			vgpu->id, reg_nonpriv, offset);
+	} else
+		intel_vgpu_default_mmio_write(vgpu, offset, p_data, bytes);
 
 	return 0;
 }
@@ -756,7 +753,7 @@ static int south_chicken2_mmio_write(struct intel_vgpu *vgpu,
 static int pri_surf_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	u32 pipe = DSPSURF_TO_PIPE(offset);
 	int event = SKL_FLIP_EVENT(pipe, PLANE_PRIMARY);
 
@@ -797,7 +794,7 @@ static int reg50080_mmio_write(struct intel_vgpu *vgpu,
 			       unsigned int offset, void *p_data,
 			       unsigned int bytes)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	enum pipe pipe = REG_50080_TO_PIPE(offset);
 	enum plane_id plane = REG_50080_TO_PLANE(offset);
 	int event = SKL_FLIP_EVENT(pipe, plane);
@@ -821,7 +818,7 @@ static int reg50080_mmio_write(struct intel_vgpu *vgpu,
 static int trigger_aux_channel_interrupt(struct intel_vgpu *vgpu,
 		unsigned int reg)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	enum intel_gvt_event_type event;
 
 	if (reg == i915_mmio_reg_offset(DP_AUX_CH_CTL(AUX_CH_A)))
@@ -836,7 +833,7 @@ static int trigger_aux_channel_interrupt(struct intel_vgpu *vgpu,
 		 reg == i915_mmio_reg_offset(DP_AUX_CH_CTL(AUX_CH_D)))
 		event = AUX_CHANNEL_D;
 	else {
-		WARN_ON(true);
+		drm_WARN_ON(&dev_priv->drm, true);
 		return -EINVAL;
 	}
 
@@ -924,11 +921,11 @@ static int dp_aux_ch_ctl_mmio_write(struct intel_vgpu *vgpu,
 	write_vreg(vgpu, offset, p_data, bytes);
 	data = vgpu_vreg(vgpu, offset);
 
-	if ((INTEL_GEN(vgpu->gvt->dev_priv) >= 9)
+	if ((INTEL_GEN(vgpu->gvt->gt->i915) >= 9)
 		&& offset != _REG_SKL_DP_AUX_CH_CTL(port_index)) {
 		/* SKL DPB/C/D aux ctl register changed */
 		return 0;
-	} else if (IS_BROADWELL(vgpu->gvt->dev_priv) &&
+	} else if (IS_BROADWELL(vgpu->gvt->gt->i915) &&
 		   offset != _REG_HSW_DP_AUX_CH_CTL(port_index)) {
 		/* write to the data registers */
 		return 0;
@@ -1244,8 +1241,7 @@ static int handle_g2v_notification(struct intel_vgpu *vgpu, int notification)
 
 static int send_display_ready_uevent(struct intel_vgpu *vgpu, int ready)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
-	struct kobject *kobj = &dev_priv->drm.primary->kdev->kobj;
+	struct kobject *kobj = &vgpu->gvt->gt->i915->drm.primary->kdev->kobj;
 	char *env[3] = {NULL, NULL, NULL};
 	char vmid_str[20];
 	char display_ready_str[20];
@@ -1306,13 +1302,15 @@ static int pvinfo_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 static int pf_write(struct intel_vgpu *vgpu,
 		unsigned int offset, void *p_data, unsigned int bytes)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	u32 val = *(u32 *)p_data;
 
 	if ((offset == _PS_1A_CTRL || offset == _PS_2A_CTRL ||
 	   offset == _PS_1B_CTRL || offset == _PS_2B_CTRL ||
 	   offset == _PS_1C_CTRL) && (val & PS_PLANE_SEL_MASK) != 0) {
-		WARN_ONCE(true, "VM(%d): guest is trying to scaling a plane\n",
-			  vgpu->id);
+		drm_WARN_ONCE(&i915->drm, true,
+			      "VM(%d): guest is trying to scaling a plane\n",
+			      vgpu->id);
 		return 0;
 	}
 
@@ -1360,13 +1358,15 @@ static int fpga_dbg_mmio_write(struct intel_vgpu *vgpu,
 static int dma_ctrl_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	u32 mode;
 
 	write_vreg(vgpu, offset, p_data, bytes);
 	mode = vgpu_vreg(vgpu, offset);
 
 	if (GFX_MODE_BIT_SET_IN_MASK(mode, START_DMA)) {
-		WARN_ONCE(1, "VM(%d): iGVT-g doesn't support GuC\n",
+		drm_WARN_ONCE(&i915->drm, 1,
+				"VM(%d): iGVT-g doesn't support GuC\n",
 				vgpu->id);
 		return 0;
 	}
@@ -1377,10 +1377,12 @@ static int dma_ctrl_write(struct intel_vgpu *vgpu, unsigned int offset,
 static int gen9_trtte_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	u32 trtte = *(u32 *)p_data;
 
 	if ((trtte & 1) && (trtte & (1 << 1)) == 0) {
-		WARN(1, "VM(%d): Use physical address for TRTT!\n",
+		drm_WARN(&i915->drm, 1,
+				"VM(%d): Use physical address for TRTT!\n",
 				vgpu->id);
 		return -EINVAL;
 	}
@@ -1427,9 +1429,9 @@ static int mailbox_write(struct intel_vgpu *vgpu, unsigned int offset,
 
 	switch (cmd) {
 	case GEN9_PCODE_READ_MEM_LATENCY:
-		if (IS_SKYLAKE(vgpu->gvt->dev_priv)
-			 || IS_KABYLAKE(vgpu->gvt->dev_priv)
-			 || IS_COFFEELAKE(vgpu->gvt->dev_priv)) {
+		if (IS_SKYLAKE(vgpu->gvt->gt->i915) ||
+		    IS_KABYLAKE(vgpu->gvt->gt->i915) ||
+		    IS_COFFEELAKE(vgpu->gvt->gt->i915)) {
 			/**
 			 * "Read memory latency" command on gen9.
 			 * Below memory latency values are read
@@ -1439,7 +1441,7 @@ static int mailbox_write(struct intel_vgpu *vgpu, unsigned int offset,
 				*data0 = 0x1e1a1100;
 			else
 				*data0 = 0x61514b3d;
-		} else if (IS_BROXTON(vgpu->gvt->dev_priv)) {
+		} else if (IS_BROXTON(vgpu->gvt->gt->i915)) {
 			/**
 			 * "Read memory latency" command on gen9.
 			 * Below memory latency values are read
@@ -1452,9 +1454,9 @@ static int mailbox_write(struct intel_vgpu *vgpu, unsigned int offset,
 		}
 		break;
 	case SKL_PCODE_CDCLK_CONTROL:
-		if (IS_SKYLAKE(vgpu->gvt->dev_priv)
-			 || IS_KABYLAKE(vgpu->gvt->dev_priv)
-			 || IS_COFFEELAKE(vgpu->gvt->dev_priv))
+		if (IS_SKYLAKE(vgpu->gvt->gt->i915) ||
+		    IS_KABYLAKE(vgpu->gvt->gt->i915) ||
+		    IS_COFFEELAKE(vgpu->gvt->gt->i915))
 			*data0 = SKL_CDCLK_READY_FOR_CHANGE;
 		break;
 	case GEN6_PCODE_READ_RC6VIDS:
@@ -1478,24 +1480,26 @@ static int hws_pga_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
 	u32 value = *(u32 *)p_data;
-	int ring_id = intel_gvt_render_mmio_to_ring_id(vgpu->gvt, offset);
+	const struct intel_engine_cs *engine =
+		intel_gvt_render_mmio_to_engine(vgpu->gvt, offset);
 
 	if (!intel_gvt_ggtt_validate_range(vgpu, value, I915_GTT_PAGE_SIZE)) {
 		gvt_vgpu_err("write invalid HWSP address, reg:0x%x, value:0x%x\n",
 			      offset, value);
 		return -EINVAL;
 	}
+
 	/*
 	 * Need to emulate all the HWSP register write to ensure host can
 	 * update the VM CSB status correctly. Here listed registers can
 	 * support BDW, SKL or other platforms with same HWSP registers.
 	 */
-	if (unlikely(ring_id < 0 || ring_id >= I915_NUM_ENGINES)) {
+	if (unlikely(!engine)) {
 		gvt_vgpu_err("access unknown hardware status page register:0x%x\n",
 			     offset);
 		return -EINVAL;
 	}
-	vgpu->hws_pga[ring_id] = value;
+	vgpu->hws_pga[engine->id] = value;
 	gvt_dbg_mmio("VM(%d) write: 0x%x to HWSP: 0x%x\n",
 		     vgpu->id, value, offset);
 
@@ -1507,7 +1511,7 @@ static int skl_power_well_ctl_write(struct intel_vgpu *vgpu,
 {
 	u32 v = *(u32 *)p_data;
 
-	if (IS_BROXTON(vgpu->gvt->dev_priv))
+	if (IS_BROXTON(vgpu->gvt->gt->i915))
 		v &= (1 << 31) | (1 << 29);
 	else
 		v &= (1 << 31) | (1 << 29) | (1 << 9) |
@@ -1654,26 +1658,24 @@ static int mmio_read_from_hw(struct intel_vgpu *vgpu,
 		unsigned int offset, void *p_data, unsigned int bytes)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
-	int ring_id;
-	u32 ring_base;
+	const struct intel_engine_cs *engine =
+		intel_gvt_render_mmio_to_engine(gvt, offset);
 
-	ring_id = intel_gvt_render_mmio_to_ring_id(gvt, offset);
 	/**
 	 * Read HW reg in following case
 	 * a. the offset isn't a ring mmio
 	 * b. the offset's ring is running on hw.
 	 * c. the offset is ring time stamp mmio
 	 */
-	if (ring_id >= 0)
-		ring_base = dev_priv->engine[ring_id]->mmio_base;
-
-	if (ring_id < 0 || vgpu  == gvt->scheduler.engine_owner[ring_id] ||
-	    offset == i915_mmio_reg_offset(RING_TIMESTAMP(ring_base)) ||
-	    offset == i915_mmio_reg_offset(RING_TIMESTAMP_UDW(ring_base))) {
-		mmio_hw_access_pre(dev_priv);
-		vgpu_vreg(vgpu, offset) = I915_READ(_MMIO(offset));
-		mmio_hw_access_post(dev_priv);
+
+	if (!engine ||
+	    vgpu == gvt->scheduler.engine_owner[engine->id] ||
+	    offset == i915_mmio_reg_offset(RING_TIMESTAMP(engine->mmio_base)) ||
+	    offset == i915_mmio_reg_offset(RING_TIMESTAMP_UDW(engine->mmio_base))) {
+		mmio_hw_access_pre(gvt->gt);
+		vgpu_vreg(vgpu, offset) =
+			intel_uncore_read(gvt->gt->uncore, _MMIO(offset));
+		mmio_hw_access_post(gvt->gt);
 	}
 
 	return intel_vgpu_default_mmio_read(vgpu, offset, p_data, bytes);
@@ -1682,22 +1684,23 @@ static int mmio_read_from_hw(struct intel_vgpu *vgpu,
 static int elsp_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
-	int ring_id = intel_gvt_render_mmio_to_ring_id(vgpu->gvt, offset);
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
+	const struct intel_engine_cs *engine = intel_gvt_render_mmio_to_engine(vgpu->gvt, offset);
 	struct intel_vgpu_execlist *execlist;
 	u32 data = *(u32 *)p_data;
 	int ret = 0;
 
-	if (WARN_ON(ring_id < 0 || ring_id >= I915_NUM_ENGINES))
+	if (drm_WARN_ON(&i915->drm, !engine))
 		return -EINVAL;
 
-	execlist = &vgpu->submission.execlist[ring_id];
+	execlist = &vgpu->submission.execlist[engine->id];
 
 	execlist->elsp_dwords.data[3 - execlist->elsp_dwords.index] = data;
 	if (execlist->elsp_dwords.index == 3) {
-		ret = intel_vgpu_submit_execlist(vgpu, ring_id);
+		ret = intel_vgpu_submit_execlist(vgpu, engine);
 		if(ret)
-			gvt_vgpu_err("fail submit workload on ring %d\n",
-				ring_id);
+			gvt_vgpu_err("fail submit workload on ring %s\n",
+				     engine->name);
 	}
 
 	++execlist->elsp_dwords.index;
@@ -1709,12 +1712,13 @@ static int ring_mode_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 		void *p_data, unsigned int bytes)
 {
 	u32 data = *(u32 *)p_data;
-	int ring_id = intel_gvt_render_mmio_to_ring_id(vgpu->gvt, offset);
+	const struct intel_engine_cs *engine =
+		intel_gvt_render_mmio_to_engine(vgpu->gvt, offset);
 	bool enable_execlist;
 	int ret;
 
 	(*(u32 *)p_data) &= ~_MASKED_BIT_ENABLE(1);
-	if (IS_COFFEELAKE(vgpu->gvt->dev_priv))
+	if (IS_COFFEELAKE(vgpu->gvt->gt->i915))
 		(*(u32 *)p_data) &= ~_MASKED_BIT_ENABLE(2);
 	write_vreg(vgpu, offset, p_data, bytes);
 
@@ -1723,7 +1727,7 @@ static int ring_mode_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 		return 0;
 	}
 
-	if (IS_COFFEELAKE(vgpu->gvt->dev_priv) &&
+	if (IS_COFFEELAKE(vgpu->gvt->gt->i915) &&
 	    data & _MASKED_BIT_ENABLE(2)) {
 		enter_failsafe_mode(vgpu, GVT_FAILSAFE_UNSUPPORTED_GUEST);
 		return 0;
@@ -1743,16 +1747,16 @@ static int ring_mode_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 			|| (data & _MASKED_BIT_DISABLE(GFX_RUN_LIST_ENABLE))) {
 		enable_execlist = !!(data & GFX_RUN_LIST_ENABLE);
 
-		gvt_dbg_core("EXECLIST %s on ring %d\n",
-				(enable_execlist ? "enabling" : "disabling"),
-				ring_id);
+		gvt_dbg_core("EXECLIST %s on ring %s\n",
+			     (enable_execlist ? "enabling" : "disabling"),
+			     engine->name);
 
 		if (!enable_execlist)
 			return 0;
 
 		ret = intel_vgpu_select_submission_ops(vgpu,
-			       BIT(ring_id),
-			       INTEL_VGPU_EXECLIST_SUBMISSION);
+						       engine->mask,
+						       INTEL_VGPU_EXECLIST_SUBMISSION);
 		if (ret)
 			return ret;
 
@@ -1876,7 +1880,7 @@ static int csfe_chicken1_mmio_write(struct intel_vgpu *vgpu,
 
 static int init_generic_mmio_info(struct intel_gvt *gvt)
 {
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct drm_i915_private *dev_priv = gvt->gt->i915;
 	int ret;
 
 	MMIO_RING_DFH(RING_IMR, D_ALL, F_CMD_ACCESS, NULL,
@@ -2693,7 +2697,7 @@ static int init_generic_mmio_info(struct intel_gvt *gvt)
 
 static int init_bdw_mmio_info(struct intel_gvt *gvt)
 {
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct drm_i915_private *dev_priv = gvt->gt->i915;
 	int ret;
 
 	MMIO_DH(GEN8_GT_IMR(0), D_BDW_PLUS, NULL, intel_vgpu_reg_imr_handler);
@@ -2882,7 +2886,7 @@ static int init_bdw_mmio_info(struct intel_gvt *gvt)
 
 static int init_skl_mmio_info(struct intel_gvt *gvt)
 {
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct drm_i915_private *dev_priv = gvt->gt->i915;
 	int ret;
 
 	MMIO_DH(FORCEWAKE_RENDER_GEN9, D_SKL_PLUS, NULL, mul_force_wake_write);
@@ -3131,7 +3135,7 @@ static int init_skl_mmio_info(struct intel_gvt *gvt)
 
 static int init_bxt_mmio_info(struct intel_gvt *gvt)
 {
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct drm_i915_private *dev_priv = gvt->gt->i915;
 	int ret;
 
 	MMIO_F(_MMIO(0x80000), 0x3000, 0, 0, 0, D_BXT, NULL, NULL);
@@ -3367,7 +3371,7 @@ static struct gvt_mmio_block mmio_blocks[] = {
 int intel_gvt_setup_mmio_info(struct intel_gvt *gvt)
 {
 	struct intel_gvt_device_info *info = &gvt->device_info;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
+	struct drm_i915_private *i915 = gvt->gt->i915;
 	int size = info->mmio_size / 4 * sizeof(*gvt->mmio.mmio_attribute);
 	int ret;
 
@@ -3379,20 +3383,20 @@ int intel_gvt_setup_mmio_info(struct intel_gvt *gvt)
 	if (ret)
 		goto err;
 
-	if (IS_BROADWELL(dev_priv)) {
+	if (IS_BROADWELL(i915)) {
 		ret = init_bdw_mmio_info(gvt);
 		if (ret)
 			goto err;
-	} else if (IS_SKYLAKE(dev_priv)
-		|| IS_KABYLAKE(dev_priv)
-		|| IS_COFFEELAKE(dev_priv)) {
+	} else if (IS_SKYLAKE(i915) ||
+		   IS_KABYLAKE(i915) ||
+		   IS_COFFEELAKE(i915)) {
 		ret = init_bdw_mmio_info(gvt);
 		if (ret)
 			goto err;
 		ret = init_skl_mmio_info(gvt);
 		if (ret)
 			goto err;
-	} else if (IS_BROXTON(dev_priv)) {
+	} else if (IS_BROXTON(i915)) {
 		ret = init_bdw_mmio_info(gvt);
 		if (ret)
 			goto err;
@@ -3541,13 +3545,14 @@ bool intel_gvt_in_force_nonpriv_whitelist(struct intel_gvt *gvt,
 int intel_vgpu_mmio_reg_rw(struct intel_vgpu *vgpu, unsigned int offset,
 			   void *pdata, unsigned int bytes, bool is_read)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_gvt *gvt = vgpu->gvt;
 	struct intel_gvt_mmio_info *mmio_info;
 	struct gvt_mmio_block *mmio_block;
 	gvt_mmio_func func;
 	int ret;
 
-	if (WARN_ON(bytes > 8))
+	if (drm_WARN_ON(&i915->drm, bytes > 8))
 		return -EINVAL;
 
 	/*
diff --git a/drivers/gpu/drm/i915/gvt/interrupt.c b/drivers/gpu/drm/i915/gvt/interrupt.c
index 11accd3e1023..540017fed908 100644
--- a/drivers/gpu/drm/i915/gvt/interrupt.c
+++ b/drivers/gpu/drm/i915/gvt/interrupt.c
@@ -245,6 +245,7 @@ int intel_vgpu_reg_ier_handler(struct intel_vgpu *vgpu,
 	unsigned int reg, void *p_data, unsigned int bytes)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
+	struct drm_i915_private *i915 = gvt->gt->i915;
 	struct intel_gvt_irq_ops *ops = gvt->irq.ops;
 	struct intel_gvt_irq_info *info;
 	u32 ier = *(u32 *)p_data;
@@ -255,7 +256,7 @@ int intel_vgpu_reg_ier_handler(struct intel_vgpu *vgpu,
 	vgpu_vreg(vgpu, reg) = ier;
 
 	info = regbase_to_irq_info(gvt, ier_to_regbase(reg));
-	if (WARN_ON(!info))
+	if (drm_WARN_ON(&i915->drm, !info))
 		return -EINVAL;
 
 	if (info->has_upstream_irq)
@@ -282,6 +283,7 @@ int intel_vgpu_reg_ier_handler(struct intel_vgpu *vgpu,
 int intel_vgpu_reg_iir_handler(struct intel_vgpu *vgpu, unsigned int reg,
 	void *p_data, unsigned int bytes)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_gvt_irq_info *info = regbase_to_irq_info(vgpu->gvt,
 		iir_to_regbase(reg));
 	u32 iir = *(u32 *)p_data;
@@ -289,7 +291,7 @@ int intel_vgpu_reg_iir_handler(struct intel_vgpu *vgpu, unsigned int reg,
 	trace_write_ir(vgpu->id, "IIR", reg, iir, vgpu_vreg(vgpu, reg),
 		       (vgpu_vreg(vgpu, reg) ^ iir));
 
-	if (WARN_ON(!info))
+	if (drm_WARN_ON(&i915->drm, !info))
 		return -EINVAL;
 
 	vgpu_vreg(vgpu, reg) &= ~iir;
@@ -319,6 +321,7 @@ static struct intel_gvt_irq_map gen8_irq_map[] = {
 static void update_upstream_irq(struct intel_vgpu *vgpu,
 		struct intel_gvt_irq_info *info)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_gvt_irq *irq = &vgpu->gvt->irq;
 	struct intel_gvt_irq_map *map = irq->irq_map;
 	struct intel_gvt_irq_info *up_irq_info = NULL;
@@ -340,7 +343,8 @@ static void update_upstream_irq(struct intel_vgpu *vgpu,
 		if (!up_irq_info)
 			up_irq_info = irq->info[map->up_irq_group];
 		else
-			WARN_ON(up_irq_info != irq->info[map->up_irq_group]);
+			drm_WARN_ON(&i915->drm, up_irq_info !=
+				    irq->info[map->up_irq_group]);
 
 		bit = map->up_irq_bit;
 
@@ -350,7 +354,7 @@ static void update_upstream_irq(struct intel_vgpu *vgpu,
 			clear_bits |= (1 << bit);
 	}
 
-	if (WARN_ON(!up_irq_info))
+	if (drm_WARN_ON(&i915->drm, !up_irq_info))
 		return;
 
 	if (up_irq_info->group == INTEL_GVT_IRQ_INFO_MASTER) {
@@ -536,7 +540,7 @@ static void gen8_init_irq(
 	SET_BIT_INFO(irq, 4, VCS_MI_FLUSH_DW, INTEL_GVT_IRQ_INFO_GT1);
 	SET_BIT_INFO(irq, 8, VCS_AS_CONTEXT_SWITCH, INTEL_GVT_IRQ_INFO_GT1);
 
-	if (HAS_ENGINE(gvt->dev_priv, VCS1)) {
+	if (HAS_ENGINE(gvt->gt->i915, VCS1)) {
 		SET_BIT_INFO(irq, 16, VCS2_MI_USER_INTERRUPT,
 			INTEL_GVT_IRQ_INFO_GT1);
 		SET_BIT_INFO(irq, 20, VCS2_MI_FLUSH_DW,
@@ -568,7 +572,7 @@ static void gen8_init_irq(
 	SET_BIT_INFO(irq, 22, DP_C_HOTPLUG, INTEL_GVT_IRQ_INFO_PCH);
 	SET_BIT_INFO(irq, 23, DP_D_HOTPLUG, INTEL_GVT_IRQ_INFO_PCH);
 
-	if (IS_BROADWELL(gvt->dev_priv)) {
+	if (IS_BROADWELL(gvt->gt->i915)) {
 		SET_BIT_INFO(irq, 25, AUX_CHANNEL_B, INTEL_GVT_IRQ_INFO_PCH);
 		SET_BIT_INFO(irq, 26, AUX_CHANNEL_C, INTEL_GVT_IRQ_INFO_PCH);
 		SET_BIT_INFO(irq, 27, AUX_CHANNEL_D, INTEL_GVT_IRQ_INFO_PCH);
@@ -581,7 +585,7 @@ static void gen8_init_irq(
 
 		SET_BIT_INFO(irq, 4, PRIMARY_C_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_C);
 		SET_BIT_INFO(irq, 5, SPRITE_C_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_C);
-	} else if (INTEL_GEN(gvt->dev_priv) >= 9) {
+	} else if (INTEL_GEN(gvt->gt->i915) >= 9) {
 		SET_BIT_INFO(irq, 25, AUX_CHANNEL_B, INTEL_GVT_IRQ_INFO_DE_PORT);
 		SET_BIT_INFO(irq, 26, AUX_CHANNEL_C, INTEL_GVT_IRQ_INFO_DE_PORT);
 		SET_BIT_INFO(irq, 27, AUX_CHANNEL_D, INTEL_GVT_IRQ_INFO_DE_PORT);
@@ -618,13 +622,14 @@ static struct intel_gvt_irq_ops gen8_irq_ops = {
 void intel_vgpu_trigger_virtual_event(struct intel_vgpu *vgpu,
 	enum intel_gvt_event_type event)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_gvt *gvt = vgpu->gvt;
 	struct intel_gvt_irq *irq = &gvt->irq;
 	gvt_event_virt_handler_t handler;
 	struct intel_gvt_irq_ops *ops = gvt->irq.ops;
 
 	handler = get_event_virt_handler(irq, event);
-	WARN_ON(!handler);
+	drm_WARN_ON(&i915->drm, !handler);
 
 	handler(irq, event, vgpu);
 
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 3259a1fa69e1..074c4efb58eb 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -108,6 +108,36 @@ struct gvt_dma {
 	struct kref ref;
 };
 
+struct kvmgt_vdev {
+	struct intel_vgpu *vgpu;
+	struct mdev_device *mdev;
+	struct vfio_region *region;
+	int num_regions;
+	struct eventfd_ctx *intx_trigger;
+	struct eventfd_ctx *msi_trigger;
+
+	/*
+	 * Two caches are used to avoid mapping duplicated pages (eg.
+	 * scratch pages). This help to reduce dma setup overhead.
+	 */
+	struct rb_root gfn_cache;
+	struct rb_root dma_addr_cache;
+	unsigned long nr_cache_entries;
+	struct mutex cache_lock;
+
+	struct notifier_block iommu_notifier;
+	struct notifier_block group_notifier;
+	struct kvm *kvm;
+	struct work_struct release_work;
+	atomic_t released;
+	struct vfio_device *vfio_device;
+};
+
+static inline struct kvmgt_vdev *kvmgt_vdev(struct intel_vgpu *vgpu)
+{
+	return intel_vgpu_vdev(vgpu);
+}
+
 static inline bool handle_valid(unsigned long handle)
 {
 	return !!(handle & ~0xff);
@@ -120,6 +150,7 @@ static bool kvmgt_guest_exit(struct kvmgt_guest_info *info);
 static void gvt_unpin_guest_page(struct intel_vgpu *vgpu, unsigned long gfn,
 		unsigned long size)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	int total_pages;
 	int npage;
 	int ret;
@@ -129,8 +160,8 @@ static void gvt_unpin_guest_page(struct intel_vgpu *vgpu, unsigned long gfn,
 	for (npage = 0; npage < total_pages; npage++) {
 		unsigned long cur_gfn = gfn + npage;
 
-		ret = vfio_unpin_pages(mdev_dev(vgpu->vdev.mdev), &cur_gfn, 1);
-		WARN_ON(ret != 1);
+		ret = vfio_unpin_pages(mdev_dev(kvmgt_vdev(vgpu)->mdev), &cur_gfn, 1);
+		drm_WARN_ON(&i915->drm, ret != 1);
 	}
 }
 
@@ -152,7 +183,7 @@ static int gvt_pin_guest_page(struct intel_vgpu *vgpu, unsigned long gfn,
 		unsigned long cur_gfn = gfn + npage;
 		unsigned long pfn;
 
-		ret = vfio_pin_pages(mdev_dev(vgpu->vdev.mdev), &cur_gfn, 1,
+		ret = vfio_pin_pages(mdev_dev(kvmgt_vdev(vgpu)->mdev), &cur_gfn, 1,
 				     IOMMU_READ | IOMMU_WRITE, &pfn);
 		if (ret != 1) {
 			gvt_vgpu_err("vfio_pin_pages failed for gfn 0x%lx, ret %d\n",
@@ -187,7 +218,7 @@ err:
 static int gvt_dma_map_page(struct intel_vgpu *vgpu, unsigned long gfn,
 		dma_addr_t *dma_addr, unsigned long size)
 {
-	struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev;
+	struct device *dev = &vgpu->gvt->gt->i915->drm.pdev->dev;
 	struct page *page = NULL;
 	int ret;
 
@@ -210,7 +241,7 @@ static int gvt_dma_map_page(struct intel_vgpu *vgpu, unsigned long gfn,
 static void gvt_dma_unmap_page(struct intel_vgpu *vgpu, unsigned long gfn,
 		dma_addr_t dma_addr, unsigned long size)
 {
-	struct device *dev = &vgpu->gvt->dev_priv->drm.pdev->dev;
+	struct device *dev = &vgpu->gvt->gt->i915->drm.pdev->dev;
 
 	dma_unmap_page(dev, dma_addr, size, PCI_DMA_BIDIRECTIONAL);
 	gvt_unpin_guest_page(vgpu, gfn, size);
@@ -219,7 +250,7 @@ static void gvt_dma_unmap_page(struct intel_vgpu *vgpu, unsigned long gfn,
 static struct gvt_dma *__gvt_cache_find_dma_addr(struct intel_vgpu *vgpu,
 		dma_addr_t dma_addr)
 {
-	struct rb_node *node = vgpu->vdev.dma_addr_cache.rb_node;
+	struct rb_node *node = kvmgt_vdev(vgpu)->dma_addr_cache.rb_node;
 	struct gvt_dma *itr;
 
 	while (node) {
@@ -237,7 +268,7 @@ static struct gvt_dma *__gvt_cache_find_dma_addr(struct intel_vgpu *vgpu,
 
 static struct gvt_dma *__gvt_cache_find_gfn(struct intel_vgpu *vgpu, gfn_t gfn)
 {
-	struct rb_node *node = vgpu->vdev.gfn_cache.rb_node;
+	struct rb_node *node = kvmgt_vdev(vgpu)->gfn_cache.rb_node;
 	struct gvt_dma *itr;
 
 	while (node) {
@@ -258,6 +289,7 @@ static int __gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn,
 {
 	struct gvt_dma *new, *itr;
 	struct rb_node **link, *parent = NULL;
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 
 	new = kzalloc(sizeof(struct gvt_dma), GFP_KERNEL);
 	if (!new)
@@ -270,7 +302,7 @@ static int __gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn,
 	kref_init(&new->ref);
 
 	/* gfn_cache maps gfn to struct gvt_dma. */
-	link = &vgpu->vdev.gfn_cache.rb_node;
+	link = &vdev->gfn_cache.rb_node;
 	while (*link) {
 		parent = *link;
 		itr = rb_entry(parent, struct gvt_dma, gfn_node);
@@ -281,11 +313,11 @@ static int __gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn,
 			link = &parent->rb_right;
 	}
 	rb_link_node(&new->gfn_node, parent, link);
-	rb_insert_color(&new->gfn_node, &vgpu->vdev.gfn_cache);
+	rb_insert_color(&new->gfn_node, &vdev->gfn_cache);
 
 	/* dma_addr_cache maps dma addr to struct gvt_dma. */
 	parent = NULL;
-	link = &vgpu->vdev.dma_addr_cache.rb_node;
+	link = &vdev->dma_addr_cache.rb_node;
 	while (*link) {
 		parent = *link;
 		itr = rb_entry(parent, struct gvt_dma, dma_addr_node);
@@ -296,46 +328,51 @@ static int __gvt_cache_add(struct intel_vgpu *vgpu, gfn_t gfn,
 			link = &parent->rb_right;
 	}
 	rb_link_node(&new->dma_addr_node, parent, link);
-	rb_insert_color(&new->dma_addr_node, &vgpu->vdev.dma_addr_cache);
+	rb_insert_color(&new->dma_addr_node, &vdev->dma_addr_cache);
 
-	vgpu->vdev.nr_cache_entries++;
+	vdev->nr_cache_entries++;
 	return 0;
 }
 
 static void __gvt_cache_remove_entry(struct intel_vgpu *vgpu,
 				struct gvt_dma *entry)
 {
-	rb_erase(&entry->gfn_node, &vgpu->vdev.gfn_cache);
-	rb_erase(&entry->dma_addr_node, &vgpu->vdev.dma_addr_cache);
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
+
+	rb_erase(&entry->gfn_node, &vdev->gfn_cache);
+	rb_erase(&entry->dma_addr_node, &vdev->dma_addr_cache);
 	kfree(entry);
-	vgpu->vdev.nr_cache_entries--;
+	vdev->nr_cache_entries--;
 }
 
 static void gvt_cache_destroy(struct intel_vgpu *vgpu)
 {
 	struct gvt_dma *dma;
 	struct rb_node *node = NULL;
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 
 	for (;;) {
-		mutex_lock(&vgpu->vdev.cache_lock);
-		node = rb_first(&vgpu->vdev.gfn_cache);
+		mutex_lock(&vdev->cache_lock);
+		node = rb_first(&vdev->gfn_cache);
 		if (!node) {
-			mutex_unlock(&vgpu->vdev.cache_lock);
+			mutex_unlock(&vdev->cache_lock);
 			break;
 		}
 		dma = rb_entry(node, struct gvt_dma, gfn_node);
 		gvt_dma_unmap_page(vgpu, dma->gfn, dma->dma_addr, dma->size);
 		__gvt_cache_remove_entry(vgpu, dma);
-		mutex_unlock(&vgpu->vdev.cache_lock);
+		mutex_unlock(&vdev->cache_lock);
 	}
 }
 
 static void gvt_cache_init(struct intel_vgpu *vgpu)
 {
-	vgpu->vdev.gfn_cache = RB_ROOT;
-	vgpu->vdev.dma_addr_cache = RB_ROOT;
-	vgpu->vdev.nr_cache_entries = 0;
-	mutex_init(&vgpu->vdev.cache_lock);
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
+
+	vdev->gfn_cache = RB_ROOT;
+	vdev->dma_addr_cache = RB_ROOT;
+	vdev->nr_cache_entries = 0;
+	mutex_init(&vdev->cache_lock);
 }
 
 static void kvmgt_protect_table_init(struct kvmgt_guest_info *info)
@@ -409,16 +446,18 @@ static void kvmgt_protect_table_del(struct kvmgt_guest_info *info,
 static size_t intel_vgpu_reg_rw_opregion(struct intel_vgpu *vgpu, char *buf,
 		size_t count, loff_t *ppos, bool iswrite)
 {
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 	unsigned int i = VFIO_PCI_OFFSET_TO_INDEX(*ppos) -
 			VFIO_PCI_NUM_REGIONS;
-	void *base = vgpu->vdev.region[i].data;
+	void *base = vdev->region[i].data;
 	loff_t pos = *ppos & VFIO_PCI_OFFSET_MASK;
 
-	if (pos >= vgpu->vdev.region[i].size || iswrite) {
+
+	if (pos >= vdev->region[i].size || iswrite) {
 		gvt_vgpu_err("invalid op or offset for Intel vgpu OpRegion\n");
 		return -EINVAL;
 	}
-	count = min(count, (size_t)(vgpu->vdev.region[i].size - pos));
+	count = min(count, (size_t)(vdev->region[i].size - pos));
 	memcpy(buf, base + pos, count);
 
 	return count;
@@ -512,7 +551,7 @@ static size_t intel_vgpu_reg_rw_edid(struct intel_vgpu *vgpu, char *buf,
 	unsigned int i = VFIO_PCI_OFFSET_TO_INDEX(*ppos) -
 			VFIO_PCI_NUM_REGIONS;
 	struct vfio_edid_region *region =
-		(struct vfio_edid_region *)vgpu->vdev.region[i].data;
+		(struct vfio_edid_region *)kvmgt_vdev(vgpu)->region[i].data;
 	loff_t pos = *ppos & VFIO_PCI_OFFSET_MASK;
 
 	if (pos < region->vfio_edid_regs.edid_offset) {
@@ -544,32 +583,34 @@ static int intel_vgpu_register_reg(struct intel_vgpu *vgpu,
 		const struct intel_vgpu_regops *ops,
 		size_t size, u32 flags, void *data)
 {
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 	struct vfio_region *region;
 
-	region = krealloc(vgpu->vdev.region,
-			(vgpu->vdev.num_regions + 1) * sizeof(*region),
+	region = krealloc(vdev->region,
+			(vdev->num_regions + 1) * sizeof(*region),
 			GFP_KERNEL);
 	if (!region)
 		return -ENOMEM;
 
-	vgpu->vdev.region = region;
-	vgpu->vdev.region[vgpu->vdev.num_regions].type = type;
-	vgpu->vdev.region[vgpu->vdev.num_regions].subtype = subtype;
-	vgpu->vdev.region[vgpu->vdev.num_regions].ops = ops;
-	vgpu->vdev.region[vgpu->vdev.num_regions].size = size;
-	vgpu->vdev.region[vgpu->vdev.num_regions].flags = flags;
-	vgpu->vdev.region[vgpu->vdev.num_regions].data = data;
-	vgpu->vdev.num_regions++;
+	vdev->region = region;
+	vdev->region[vdev->num_regions].type = type;
+	vdev->region[vdev->num_regions].subtype = subtype;
+	vdev->region[vdev->num_regions].ops = ops;
+	vdev->region[vdev->num_regions].size = size;
+	vdev->region[vdev->num_regions].flags = flags;
+	vdev->region[vdev->num_regions].data = data;
+	vdev->num_regions++;
 	return 0;
 }
 
 static int kvmgt_get_vfio_device(void *p_vgpu)
 {
 	struct intel_vgpu *vgpu = (struct intel_vgpu *)p_vgpu;
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 
-	vgpu->vdev.vfio_device = vfio_device_get_from_dev(
-		mdev_dev(vgpu->vdev.mdev));
-	if (!vgpu->vdev.vfio_device) {
+	vdev->vfio_device = vfio_device_get_from_dev(
+		mdev_dev(vdev->mdev));
+	if (!vdev->vfio_device) {
 		gvt_vgpu_err("failed to get vfio device\n");
 		return -ENODEV;
 	}
@@ -637,10 +678,12 @@ static int kvmgt_set_edid(void *p_vgpu, int port_num)
 
 static void kvmgt_put_vfio_device(void *vgpu)
 {
-	if (WARN_ON(!((struct intel_vgpu *)vgpu)->vdev.vfio_device))
+	struct kvmgt_vdev *vdev = kvmgt_vdev((struct intel_vgpu *)vgpu);
+
+	if (WARN_ON(!vdev->vfio_device))
 		return;
 
-	vfio_device_put(((struct intel_vgpu *)vgpu)->vdev.vfio_device);
+	vfio_device_put(vdev->vfio_device);
 }
 
 static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
@@ -669,9 +712,9 @@ static int intel_vgpu_create(struct kobject *kobj, struct mdev_device *mdev)
 		goto out;
 	}
 
-	INIT_WORK(&vgpu->vdev.release_work, intel_vgpu_release_work);
+	INIT_WORK(&kvmgt_vdev(vgpu)->release_work, intel_vgpu_release_work);
 
-	vgpu->vdev.mdev = mdev;
+	kvmgt_vdev(vgpu)->mdev = mdev;
 	mdev_set_drvdata(mdev, vgpu);
 
 	gvt_dbg_core("intel_vgpu_create succeeded for mdev: %s\n",
@@ -696,9 +739,10 @@ static int intel_vgpu_remove(struct mdev_device *mdev)
 static int intel_vgpu_iommu_notifier(struct notifier_block *nb,
 				     unsigned long action, void *data)
 {
-	struct intel_vgpu *vgpu = container_of(nb,
-					struct intel_vgpu,
-					vdev.iommu_notifier);
+	struct kvmgt_vdev *vdev = container_of(nb,
+					       struct kvmgt_vdev,
+					       iommu_notifier);
+	struct intel_vgpu *vgpu = vdev->vgpu;
 
 	if (action == VFIO_IOMMU_NOTIFY_DMA_UNMAP) {
 		struct vfio_iommu_type1_dma_unmap *unmap = data;
@@ -708,7 +752,7 @@ static int intel_vgpu_iommu_notifier(struct notifier_block *nb,
 		iov_pfn = unmap->iova >> PAGE_SHIFT;
 		end_iov_pfn = iov_pfn + unmap->size / PAGE_SIZE;
 
-		mutex_lock(&vgpu->vdev.cache_lock);
+		mutex_lock(&vdev->cache_lock);
 		for (; iov_pfn < end_iov_pfn; iov_pfn++) {
 			entry = __gvt_cache_find_gfn(vgpu, iov_pfn);
 			if (!entry)
@@ -718,7 +762,7 @@ static int intel_vgpu_iommu_notifier(struct notifier_block *nb,
 					   entry->size);
 			__gvt_cache_remove_entry(vgpu, entry);
 		}
-		mutex_unlock(&vgpu->vdev.cache_lock);
+		mutex_unlock(&vdev->cache_lock);
 	}
 
 	return NOTIFY_OK;
@@ -727,16 +771,16 @@ static int intel_vgpu_iommu_notifier(struct notifier_block *nb,
 static int intel_vgpu_group_notifier(struct notifier_block *nb,
 				     unsigned long action, void *data)
 {
-	struct intel_vgpu *vgpu = container_of(nb,
-					struct intel_vgpu,
-					vdev.group_notifier);
+	struct kvmgt_vdev *vdev = container_of(nb,
+					       struct kvmgt_vdev,
+					       group_notifier);
 
 	/* the only action we care about */
 	if (action == VFIO_GROUP_NOTIFY_SET_KVM) {
-		vgpu->vdev.kvm = data;
+		vdev->kvm = data;
 
 		if (!data)
-			schedule_work(&vgpu->vdev.release_work);
+			schedule_work(&vdev->release_work);
 	}
 
 	return NOTIFY_OK;
@@ -745,15 +789,16 @@ static int intel_vgpu_group_notifier(struct notifier_block *nb,
 static int intel_vgpu_open(struct mdev_device *mdev)
 {
 	struct intel_vgpu *vgpu = mdev_get_drvdata(mdev);
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 	unsigned long events;
 	int ret;
 
-	vgpu->vdev.iommu_notifier.notifier_call = intel_vgpu_iommu_notifier;
-	vgpu->vdev.group_notifier.notifier_call = intel_vgpu_group_notifier;
+	vdev->iommu_notifier.notifier_call = intel_vgpu_iommu_notifier;
+	vdev->group_notifier.notifier_call = intel_vgpu_group_notifier;
 
 	events = VFIO_IOMMU_NOTIFY_DMA_UNMAP;
 	ret = vfio_register_notifier(mdev_dev(mdev), VFIO_IOMMU_NOTIFY, &events,
-				&vgpu->vdev.iommu_notifier);
+				&vdev->iommu_notifier);
 	if (ret != 0) {
 		gvt_vgpu_err("vfio_register_notifier for iommu failed: %d\n",
 			ret);
@@ -762,7 +807,7 @@ static int intel_vgpu_open(struct mdev_device *mdev)
 
 	events = VFIO_GROUP_NOTIFY_SET_KVM;
 	ret = vfio_register_notifier(mdev_dev(mdev), VFIO_GROUP_NOTIFY, &events,
-				&vgpu->vdev.group_notifier);
+				&vdev->group_notifier);
 	if (ret != 0) {
 		gvt_vgpu_err("vfio_register_notifier for group failed: %d\n",
 			ret);
@@ -781,51 +826,56 @@ static int intel_vgpu_open(struct mdev_device *mdev)
 
 	intel_gvt_ops->vgpu_activate(vgpu);
 
-	atomic_set(&vgpu->vdev.released, 0);
+	atomic_set(&vdev->released, 0);
 	return ret;
 
 undo_group:
 	vfio_unregister_notifier(mdev_dev(mdev), VFIO_GROUP_NOTIFY,
-					&vgpu->vdev.group_notifier);
+					&vdev->group_notifier);
 
 undo_iommu:
 	vfio_unregister_notifier(mdev_dev(mdev), VFIO_IOMMU_NOTIFY,
-					&vgpu->vdev.iommu_notifier);
+					&vdev->iommu_notifier);
 out:
 	return ret;
 }
 
 static void intel_vgpu_release_msi_eventfd_ctx(struct intel_vgpu *vgpu)
 {
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 	struct eventfd_ctx *trigger;
 
-	trigger = vgpu->vdev.msi_trigger;
+	trigger = vdev->msi_trigger;
 	if (trigger) {
 		eventfd_ctx_put(trigger);
-		vgpu->vdev.msi_trigger = NULL;
+		vdev->msi_trigger = NULL;
 	}
 }
 
 static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 {
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct kvmgt_guest_info *info;
 	int ret;
 
 	if (!handle_valid(vgpu->handle))
 		return;
 
-	if (atomic_cmpxchg(&vgpu->vdev.released, 0, 1))
+	if (atomic_cmpxchg(&vdev->released, 0, 1))
 		return;
 
 	intel_gvt_ops->vgpu_release(vgpu);
 
-	ret = vfio_unregister_notifier(mdev_dev(vgpu->vdev.mdev), VFIO_IOMMU_NOTIFY,
-					&vgpu->vdev.iommu_notifier);
-	WARN(ret, "vfio_unregister_notifier for iommu failed: %d\n", ret);
+	ret = vfio_unregister_notifier(mdev_dev(vdev->mdev), VFIO_IOMMU_NOTIFY,
+					&vdev->iommu_notifier);
+	drm_WARN(&i915->drm, ret,
+		 "vfio_unregister_notifier for iommu failed: %d\n", ret);
 
-	ret = vfio_unregister_notifier(mdev_dev(vgpu->vdev.mdev), VFIO_GROUP_NOTIFY,
-					&vgpu->vdev.group_notifier);
-	WARN(ret, "vfio_unregister_notifier for group failed: %d\n", ret);
+	ret = vfio_unregister_notifier(mdev_dev(vdev->mdev), VFIO_GROUP_NOTIFY,
+					&vdev->group_notifier);
+	drm_WARN(&i915->drm, ret,
+		 "vfio_unregister_notifier for group failed: %d\n", ret);
 
 	/* dereference module reference taken at open */
 	module_put(THIS_MODULE);
@@ -835,7 +885,7 @@ static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 
 	intel_vgpu_release_msi_eventfd_ctx(vgpu);
 
-	vgpu->vdev.kvm = NULL;
+	vdev->kvm = NULL;
 	vgpu->handle = 0;
 }
 
@@ -848,10 +898,10 @@ static void intel_vgpu_release(struct mdev_device *mdev)
 
 static void intel_vgpu_release_work(struct work_struct *work)
 {
-	struct intel_vgpu *vgpu = container_of(work, struct intel_vgpu,
-					vdev.release_work);
+	struct kvmgt_vdev *vdev = container_of(work, struct kvmgt_vdev,
+					       release_work);
 
-	__intel_vgpu_release(vgpu);
+	__intel_vgpu_release(vdev->vgpu);
 }
 
 static u64 intel_vgpu_get_bar_addr(struct intel_vgpu *vgpu, int bar)
@@ -913,7 +963,7 @@ static int intel_vgpu_aperture_rw(struct intel_vgpu *vgpu, u64 off,
 		return -EINVAL;
 	}
 
-	aperture_va = io_mapping_map_wc(&vgpu->gvt->dev_priv->ggtt.iomap,
+	aperture_va = io_mapping_map_wc(&vgpu->gvt->gt->ggtt->iomap,
 					ALIGN_DOWN(off, PAGE_SIZE),
 					count + offset_in_page(off));
 	if (!aperture_va)
@@ -933,12 +983,13 @@ static ssize_t intel_vgpu_rw(struct mdev_device *mdev, char *buf,
 			size_t count, loff_t *ppos, bool is_write)
 {
 	struct intel_vgpu *vgpu = mdev_get_drvdata(mdev);
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 	unsigned int index = VFIO_PCI_OFFSET_TO_INDEX(*ppos);
 	u64 pos = *ppos & VFIO_PCI_OFFSET_MASK;
 	int ret = -EINVAL;
 
 
-	if (index >= VFIO_PCI_NUM_REGIONS + vgpu->vdev.num_regions) {
+	if (index >= VFIO_PCI_NUM_REGIONS + vdev->num_regions) {
 		gvt_vgpu_err("invalid index: %u\n", index);
 		return -EINVAL;
 	}
@@ -967,11 +1018,11 @@ static ssize_t intel_vgpu_rw(struct mdev_device *mdev, char *buf,
 	case VFIO_PCI_ROM_REGION_INDEX:
 		break;
 	default:
-		if (index >= VFIO_PCI_NUM_REGIONS + vgpu->vdev.num_regions)
+		if (index >= VFIO_PCI_NUM_REGIONS + vdev->num_regions)
 			return -EINVAL;
 
 		index -= VFIO_PCI_NUM_REGIONS;
-		return vgpu->vdev.region[index].ops->rw(vgpu, buf, count,
+		return vdev->region[index].ops->rw(vgpu, buf, count,
 				ppos, is_write);
 	}
 
@@ -1224,7 +1275,7 @@ static int intel_vgpu_set_msi_trigger(struct intel_vgpu *vgpu,
 			gvt_vgpu_err("eventfd_ctx_fdget failed\n");
 			return PTR_ERR(trigger);
 		}
-		vgpu->vdev.msi_trigger = trigger;
+		kvmgt_vdev(vgpu)->msi_trigger = trigger;
 	} else if ((flags & VFIO_IRQ_SET_DATA_NONE) && !count)
 		intel_vgpu_release_msi_eventfd_ctx(vgpu);
 
@@ -1276,6 +1327,7 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd,
 			     unsigned long arg)
 {
 	struct intel_vgpu *vgpu = mdev_get_drvdata(mdev);
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 	unsigned long minsz;
 
 	gvt_dbg_core("vgpu%d ioctl, cmd: %d\n", vgpu->id, cmd);
@@ -1294,7 +1346,7 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd,
 		info.flags = VFIO_DEVICE_FLAGS_PCI;
 		info.flags |= VFIO_DEVICE_FLAGS_RESET;
 		info.num_regions = VFIO_PCI_NUM_REGIONS +
-				vgpu->vdev.num_regions;
+				vdev->num_regions;
 		info.num_irqs = VFIO_PCI_NUM_IRQS;
 
 		return copy_to_user((void __user *)arg, &info, minsz) ?
@@ -1385,22 +1437,22 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd,
 					.header.version = 1 };
 
 				if (info.index >= VFIO_PCI_NUM_REGIONS +
-						vgpu->vdev.num_regions)
+						vdev->num_regions)
 					return -EINVAL;
 				info.index =
 					array_index_nospec(info.index,
 							VFIO_PCI_NUM_REGIONS +
-							vgpu->vdev.num_regions);
+							vdev->num_regions);
 
 				i = info.index - VFIO_PCI_NUM_REGIONS;
 
 				info.offset =
 					VFIO_PCI_INDEX_TO_OFFSET(info.index);
-				info.size = vgpu->vdev.region[i].size;
-				info.flags = vgpu->vdev.region[i].flags;
+				info.size = vdev->region[i].size;
+				info.flags = vdev->region[i].flags;
 
-				cap_type.type = vgpu->vdev.region[i].type;
-				cap_type.subtype = vgpu->vdev.region[i].subtype;
+				cap_type.type = vdev->region[i].type;
+				cap_type.subtype = vdev->region[i].subtype;
 
 				ret = vfio_info_add_capability(&caps,
 							&cap_type.header,
@@ -1597,12 +1649,10 @@ static struct mdev_parent_ops intel_vgpu_ops = {
 
 static int kvmgt_host_init(struct device *dev, void *gvt, const void *ops)
 {
-	struct attribute **kvm_type_attrs;
 	struct attribute_group **kvm_vgpu_type_groups;
 
 	intel_gvt_ops = ops;
-	if (!intel_gvt_ops->get_gvt_attrs(&kvm_type_attrs,
-			&kvm_vgpu_type_groups))
+	if (!intel_gvt_ops->get_gvt_attrs(&kvm_vgpu_type_groups))
 		return -EFAULT;
 	intel_vgpu_ops.supported_type_groups = kvm_vgpu_type_groups;
 
@@ -1742,13 +1792,15 @@ static int kvmgt_guest_init(struct mdev_device *mdev)
 {
 	struct kvmgt_guest_info *info;
 	struct intel_vgpu *vgpu;
+	struct kvmgt_vdev *vdev;
 	struct kvm *kvm;
 
 	vgpu = mdev_get_drvdata(mdev);
 	if (handle_valid(vgpu->handle))
 		return -EEXIST;
 
-	kvm = vgpu->vdev.kvm;
+	vdev = kvmgt_vdev(vgpu);
+	kvm = vdev->kvm;
 	if (!kvm || kvm->mm != current->mm) {
 		gvt_vgpu_err("KVM is required to use Intel vGPU\n");
 		return -ESRCH;
@@ -1769,8 +1821,6 @@ static int kvmgt_guest_init(struct mdev_device *mdev)
 	kvmgt_protect_table_init(info);
 	gvt_cache_init(vgpu);
 
-	init_completion(&vgpu->vblank_done);
-
 	info->track_node.track_write = kvmgt_page_track_write;
 	info->track_node.track_flush_slot = kvmgt_page_track_flush_slot;
 	kvm_page_track_register_notifier(kvm, &info->track_node);
@@ -1778,7 +1828,7 @@ static int kvmgt_guest_init(struct mdev_device *mdev)
 	info->debugfs_cache_entries = debugfs_create_ulong(
 						"kvmgt_nr_cache_entries",
 						0444, vgpu->debugfs,
-						&vgpu->vdev.nr_cache_entries);
+						&vdev->nr_cache_entries);
 	return 0;
 }
 
@@ -1795,9 +1845,17 @@ static bool kvmgt_guest_exit(struct kvmgt_guest_info *info)
 	return true;
 }
 
-static int kvmgt_attach_vgpu(void *vgpu, unsigned long *handle)
+static int kvmgt_attach_vgpu(void *p_vgpu, unsigned long *handle)
 {
-	/* nothing to do here */
+	struct intel_vgpu *vgpu = (struct intel_vgpu *)p_vgpu;
+
+	vgpu->vdev = kzalloc(sizeof(struct kvmgt_vdev), GFP_KERNEL);
+
+	if (!vgpu->vdev)
+		return -ENOMEM;
+
+	kvmgt_vdev(vgpu)->vgpu = vgpu;
+
 	return 0;
 }
 
@@ -1805,29 +1863,34 @@ static void kvmgt_detach_vgpu(void *p_vgpu)
 {
 	int i;
 	struct intel_vgpu *vgpu = (struct intel_vgpu *)p_vgpu;
+	struct kvmgt_vdev *vdev = kvmgt_vdev(vgpu);
 
-	if (!vgpu->vdev.region)
+	if (!vdev->region)
 		return;
 
-	for (i = 0; i < vgpu->vdev.num_regions; i++)
-		if (vgpu->vdev.region[i].ops->release)
-			vgpu->vdev.region[i].ops->release(vgpu,
-					&vgpu->vdev.region[i]);
-	vgpu->vdev.num_regions = 0;
-	kfree(vgpu->vdev.region);
-	vgpu->vdev.region = NULL;
+	for (i = 0; i < vdev->num_regions; i++)
+		if (vdev->region[i].ops->release)
+			vdev->region[i].ops->release(vgpu,
+					&vdev->region[i]);
+	vdev->num_regions = 0;
+	kfree(vdev->region);
+	vdev->region = NULL;
+
+	kfree(vdev);
 }
 
 static int kvmgt_inject_msi(unsigned long handle, u32 addr, u16 data)
 {
 	struct kvmgt_guest_info *info;
 	struct intel_vgpu *vgpu;
+	struct kvmgt_vdev *vdev;
 
 	if (!handle_valid(handle))
 		return -ESRCH;
 
 	info = (struct kvmgt_guest_info *)handle;
 	vgpu = info->vgpu;
+	vdev = kvmgt_vdev(vgpu);
 
 	/*
 	 * When guest is poweroff, msi_trigger is set to NULL, but vgpu's
@@ -1838,10 +1901,10 @@ static int kvmgt_inject_msi(unsigned long handle, u32 addr, u16 data)
 	 * enabled by guest. so if msi_trigger is null, success is still
 	 * returned and don't inject interrupt into guest.
 	 */
-	if (vgpu->vdev.msi_trigger == NULL)
+	if (vdev->msi_trigger == NULL)
 		return 0;
 
-	if (eventfd_signal(vgpu->vdev.msi_trigger, 1) == 1)
+	if (eventfd_signal(vdev->msi_trigger, 1) == 1)
 		return 0;
 
 	return -EFAULT;
@@ -1867,26 +1930,26 @@ static unsigned long kvmgt_gfn_to_pfn(unsigned long handle, unsigned long gfn)
 static int kvmgt_dma_map_guest_page(unsigned long handle, unsigned long gfn,
 		unsigned long size, dma_addr_t *dma_addr)
 {
-	struct kvmgt_guest_info *info;
 	struct intel_vgpu *vgpu;
+	struct kvmgt_vdev *vdev;
 	struct gvt_dma *entry;
 	int ret;
 
 	if (!handle_valid(handle))
 		return -EINVAL;
 
-	info = (struct kvmgt_guest_info *)handle;
-	vgpu = info->vgpu;
+	vgpu = ((struct kvmgt_guest_info *)handle)->vgpu;
+	vdev = kvmgt_vdev(vgpu);
 
-	mutex_lock(&info->vgpu->vdev.cache_lock);
+	mutex_lock(&vdev->cache_lock);
 
-	entry = __gvt_cache_find_gfn(info->vgpu, gfn);
+	entry = __gvt_cache_find_gfn(vgpu, gfn);
 	if (!entry) {
 		ret = gvt_dma_map_page(vgpu, gfn, dma_addr, size);
 		if (ret)
 			goto err_unlock;
 
-		ret = __gvt_cache_add(info->vgpu, gfn, *dma_addr, size);
+		ret = __gvt_cache_add(vgpu, gfn, *dma_addr, size);
 		if (ret)
 			goto err_unmap;
 	} else if (entry->size != size) {
@@ -1898,7 +1961,7 @@ static int kvmgt_dma_map_guest_page(unsigned long handle, unsigned long gfn,
 		if (ret)
 			goto err_unlock;
 
-		ret = __gvt_cache_add(info->vgpu, gfn, *dma_addr, size);
+		ret = __gvt_cache_add(vgpu, gfn, *dma_addr, size);
 		if (ret)
 			goto err_unmap;
 	} else {
@@ -1906,19 +1969,20 @@ static int kvmgt_dma_map_guest_page(unsigned long handle, unsigned long gfn,
 		*dma_addr = entry->dma_addr;
 	}
 
-	mutex_unlock(&info->vgpu->vdev.cache_lock);
+	mutex_unlock(&vdev->cache_lock);
 	return 0;
 
 err_unmap:
 	gvt_dma_unmap_page(vgpu, gfn, *dma_addr, size);
 err_unlock:
-	mutex_unlock(&info->vgpu->vdev.cache_lock);
+	mutex_unlock(&vdev->cache_lock);
 	return ret;
 }
 
 static int kvmgt_dma_pin_guest_page(unsigned long handle, dma_addr_t dma_addr)
 {
 	struct kvmgt_guest_info *info;
+	struct kvmgt_vdev *vdev;
 	struct gvt_dma *entry;
 	int ret = 0;
 
@@ -1926,14 +1990,15 @@ static int kvmgt_dma_pin_guest_page(unsigned long handle, dma_addr_t dma_addr)
 		return -ENODEV;
 
 	info = (struct kvmgt_guest_info *)handle;
+	vdev = kvmgt_vdev(info->vgpu);
 
-	mutex_lock(&info->vgpu->vdev.cache_lock);
+	mutex_lock(&vdev->cache_lock);
 	entry = __gvt_cache_find_dma_addr(info->vgpu, dma_addr);
 	if (entry)
 		kref_get(&entry->ref);
 	else
 		ret = -ENOMEM;
-	mutex_unlock(&info->vgpu->vdev.cache_lock);
+	mutex_unlock(&vdev->cache_lock);
 
 	return ret;
 }
@@ -1949,19 +2014,21 @@ static void __gvt_dma_release(struct kref *ref)
 
 static void kvmgt_dma_unmap_guest_page(unsigned long handle, dma_addr_t dma_addr)
 {
-	struct kvmgt_guest_info *info;
+	struct intel_vgpu *vgpu;
+	struct kvmgt_vdev *vdev;
 	struct gvt_dma *entry;
 
 	if (!handle_valid(handle))
 		return;
 
-	info = (struct kvmgt_guest_info *)handle;
+	vgpu = ((struct kvmgt_guest_info *)handle)->vgpu;
+	vdev = kvmgt_vdev(vgpu);
 
-	mutex_lock(&info->vgpu->vdev.cache_lock);
-	entry = __gvt_cache_find_dma_addr(info->vgpu, dma_addr);
+	mutex_lock(&vdev->cache_lock);
+	entry = __gvt_cache_find_dma_addr(vgpu, dma_addr);
 	if (entry)
 		kref_put(&entry->ref, __gvt_dma_release);
-	mutex_unlock(&info->vgpu->vdev.cache_lock);
+	mutex_unlock(&vdev->cache_lock);
 }
 
 static int kvmgt_rw_gpa(unsigned long handle, unsigned long gpa,
diff --git a/drivers/gpu/drm/i915/gvt/mmio.c b/drivers/gpu/drm/i915/gvt/mmio.c
index a55178884d67..291993615af9 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.c
+++ b/drivers/gpu/drm/i915/gvt/mmio.c
@@ -103,6 +103,7 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, u64 pa,
 		void *p_data, unsigned int bytes)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
+	struct drm_i915_private *i915 = gvt->gt->i915;
 	unsigned int offset = 0;
 	int ret = -EINVAL;
 
@@ -114,15 +115,17 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, u64 pa,
 
 	offset = intel_vgpu_gpa_to_mmio_offset(vgpu, pa);
 
-	if (WARN_ON(bytes > 8))
+	if (drm_WARN_ON(&i915->drm, bytes > 8))
 		goto err;
 
 	if (reg_is_gtt(gvt, offset)) {
-		if (WARN_ON(!IS_ALIGNED(offset, 4) && !IS_ALIGNED(offset, 8)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, 4) &&
+				!IS_ALIGNED(offset, 8)))
 			goto err;
-		if (WARN_ON(bytes != 4 && bytes != 8))
+		if (drm_WARN_ON(&i915->drm, bytes != 4 && bytes != 8))
 			goto err;
-		if (WARN_ON(!reg_is_gtt(gvt, offset + bytes - 1)))
+		if (drm_WARN_ON(&i915->drm,
+				!reg_is_gtt(gvt, offset + bytes - 1)))
 			goto err;
 
 		ret = intel_vgpu_emulate_ggtt_mmio_read(vgpu, offset,
@@ -132,16 +135,16 @@ int intel_vgpu_emulate_mmio_read(struct intel_vgpu *vgpu, u64 pa,
 		goto out;
 	}
 
-	if (WARN_ON_ONCE(!reg_is_mmio(gvt, offset))) {
+	if (drm_WARN_ON_ONCE(&i915->drm, !reg_is_mmio(gvt, offset))) {
 		ret = intel_gvt_hypervisor_read_gpa(vgpu, pa, p_data, bytes);
 		goto out;
 	}
 
-	if (WARN_ON(!reg_is_mmio(gvt, offset + bytes - 1)))
+	if (drm_WARN_ON(&i915->drm, !reg_is_mmio(gvt, offset + bytes - 1)))
 		goto err;
 
 	if (!intel_gvt_mmio_is_unalign(gvt, offset)) {
-		if (WARN_ON(!IS_ALIGNED(offset, bytes)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, bytes)))
 			goto err;
 	}
 
@@ -175,6 +178,7 @@ int intel_vgpu_emulate_mmio_write(struct intel_vgpu *vgpu, u64 pa,
 		void *p_data, unsigned int bytes)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
+	struct drm_i915_private *i915 = gvt->gt->i915;
 	unsigned int offset = 0;
 	int ret = -EINVAL;
 
@@ -187,15 +191,17 @@ int intel_vgpu_emulate_mmio_write(struct intel_vgpu *vgpu, u64 pa,
 
 	offset = intel_vgpu_gpa_to_mmio_offset(vgpu, pa);
 
-	if (WARN_ON(bytes > 8))
+	if (drm_WARN_ON(&i915->drm, bytes > 8))
 		goto err;
 
 	if (reg_is_gtt(gvt, offset)) {
-		if (WARN_ON(!IS_ALIGNED(offset, 4) && !IS_ALIGNED(offset, 8)))
+		if (drm_WARN_ON(&i915->drm, !IS_ALIGNED(offset, 4) &&
+				!IS_ALIGNED(offset, 8)))
 			goto err;
-		if (WARN_ON(bytes != 4 && bytes != 8))
+		if (drm_WARN_ON(&i915->drm, bytes != 4 && bytes != 8))
 			goto err;
-		if (WARN_ON(!reg_is_gtt(gvt, offset + bytes - 1)))
+		if (drm_WARN_ON(&i915->drm,
+				!reg_is_gtt(gvt, offset + bytes - 1)))
 			goto err;
 
 		ret = intel_vgpu_emulate_ggtt_mmio_write(vgpu, offset,
@@ -205,7 +211,7 @@ int intel_vgpu_emulate_mmio_write(struct intel_vgpu *vgpu, u64 pa,
 		goto out;
 	}
 
-	if (WARN_ON_ONCE(!reg_is_mmio(gvt, offset))) {
+	if (drm_WARN_ON_ONCE(&i915->drm, !reg_is_mmio(gvt, offset))) {
 		ret = intel_gvt_hypervisor_write_gpa(vgpu, pa, p_data, bytes);
 		goto out;
 	}
@@ -245,7 +251,7 @@ void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu, bool dmlr)
 		/* set the bit 0:2(Core C-State ) to C0 */
 		vgpu_vreg_t(vgpu, GEN6_GT_CORE_STATUS) = 0;
 
-		if (IS_BROXTON(vgpu->gvt->dev_priv)) {
+		if (IS_BROXTON(vgpu->gvt->gt->i915)) {
 			vgpu_vreg_t(vgpu, BXT_P_CR_GT_DISP_PWRON) &=
 				    ~(BIT(0) | BIT(1));
 			vgpu_vreg_t(vgpu, BXT_PORT_CL1CM_DW0(DPIO_PHY0)) &=
diff --git a/drivers/gpu/drm/i915/gvt/mmio.h b/drivers/gpu/drm/i915/gvt/mmio.h
index 2e68f4b02c94..cc4812648bf4 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.h
+++ b/drivers/gpu/drm/i915/gvt/mmio.h
@@ -69,8 +69,8 @@ struct intel_gvt_mmio_info {
 	struct hlist_node node;
 };
 
-int intel_gvt_render_mmio_to_ring_id(struct intel_gvt *gvt,
-		unsigned int reg);
+const struct intel_engine_cs *
+intel_gvt_render_mmio_to_engine(struct intel_gvt *gvt, unsigned int reg);
 unsigned long intel_gvt_get_device_type(struct intel_gvt *gvt);
 bool intel_gvt_match_device(struct intel_gvt *gvt, unsigned long device);
 
diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.c b/drivers/gpu/drm/i915/gvt/mmio_context.c
index aaf15916d29a..2ccaf78f96e8 100644
--- a/drivers/gpu/drm/i915/gvt/mmio_context.c
+++ b/drivers/gpu/drm/i915/gvt/mmio_context.c
@@ -157,12 +157,13 @@ static u32 gen9_mocs_mmio_offset_list[] = {
 	[VECS0] = 0xcb00,
 };
 
-static void load_render_mocs(struct drm_i915_private *dev_priv)
+static void load_render_mocs(const struct intel_engine_cs *engine)
 {
-	struct intel_gvt *gvt = dev_priv->gvt;
-	i915_reg_t offset;
+	struct intel_gvt *gvt = engine->i915->gvt;
+	struct intel_uncore *uncore = engine->uncore;
 	u32 cnt = gvt->engine_mmio_list.mocs_mmio_offset_list_cnt;
 	u32 *regs = gvt->engine_mmio_list.mocs_mmio_offset_list;
+	i915_reg_t offset;
 	int ring_id, i;
 
 	/* Platform doesn't have mocs mmios. */
@@ -170,12 +171,13 @@ static void load_render_mocs(struct drm_i915_private *dev_priv)
 		return;
 
 	for (ring_id = 0; ring_id < cnt; ring_id++) {
-		if (!HAS_ENGINE(dev_priv, ring_id))
+		if (!HAS_ENGINE(engine->i915, ring_id))
 			continue;
+
 		offset.reg = regs[ring_id];
 		for (i = 0; i < GEN9_MOCS_SIZE; i++) {
 			gen9_render_mocs.control_table[ring_id][i] =
-				I915_READ_FW(offset);
+				intel_uncore_read_fw(uncore, offset);
 			offset.reg += 4;
 		}
 	}
@@ -183,7 +185,7 @@ static void load_render_mocs(struct drm_i915_private *dev_priv)
 	offset.reg = 0xb020;
 	for (i = 0; i < GEN9_MOCS_SIZE / 2; i++) {
 		gen9_render_mocs.l3cc_table[i] =
-			I915_READ_FW(offset);
+			intel_uncore_read_fw(uncore, offset);
 		offset.reg += 4;
 	}
 	gen9_render_mocs.initialized = true;
@@ -214,13 +216,11 @@ restore_context_mmio_for_inhibit(struct intel_vgpu *vgpu,
 	*cs++ = MI_LOAD_REGISTER_IMM(count);
 	for (mmio = gvt->engine_mmio_list.mmio;
 	     i915_mmio_reg_valid(mmio->reg); mmio++) {
-		if (mmio->ring_id != ring_id ||
-		    !mmio->in_context)
+		if (mmio->id != ring_id || !mmio->in_context)
 			continue;
 
 		*cs++ = i915_mmio_reg_offset(mmio->reg);
-		*cs++ = vgpu_vreg_t(vgpu, mmio->reg) |
-				(mmio->mask << 16);
+		*cs++ = vgpu_vreg_t(vgpu, mmio->reg) | (mmio->mask << 16);
 		gvt_dbg_core("add lri reg pair 0x%x:0x%x in inhibit ctx, vgpu:%d, rind_id:%d\n",
 			      *(cs-2), *(cs-1), vgpu->id, ring_id);
 	}
@@ -344,10 +344,10 @@ static u32 gen8_tlb_mmio_offset_list[] = {
 	[VECS0] = 0x4270,
 };
 
-static void handle_tlb_pending_event(struct intel_vgpu *vgpu, int ring_id)
+static void handle_tlb_pending_event(struct intel_vgpu *vgpu,
+				     const struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
-	struct intel_uncore *uncore = &dev_priv->uncore;
+	struct intel_uncore *uncore = engine->uncore;
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	u32 *regs = vgpu->gvt->engine_mmio_list.tlb_mmio_offset_list;
 	u32 cnt = vgpu->gvt->engine_mmio_list.tlb_mmio_offset_list_cnt;
@@ -357,13 +357,13 @@ static void handle_tlb_pending_event(struct intel_vgpu *vgpu, int ring_id)
 	if (!regs)
 		return;
 
-	if (WARN_ON(ring_id >= cnt))
+	if (drm_WARN_ON(&engine->i915->drm, engine->id >= cnt))
 		return;
 
-	if (!test_and_clear_bit(ring_id, (void *)s->tlb_handle_pending))
+	if (!test_and_clear_bit(engine->id, (void *)s->tlb_handle_pending))
 		return;
 
-	reg = _MMIO(regs[ring_id]);
+	reg = _MMIO(regs[engine->id]);
 
 	/* WaForceWakeRenderDuringMmioTLBInvalidate:skl
 	 * we need to put a forcewake when invalidating RCS TLB caches,
@@ -372,30 +372,27 @@ static void handle_tlb_pending_event(struct intel_vgpu *vgpu, int ring_id)
 	 */
 	fw = intel_uncore_forcewake_for_reg(uncore, reg,
 					    FW_REG_READ | FW_REG_WRITE);
-	if (ring_id == RCS0 && INTEL_GEN(dev_priv) >= 9)
+	if (engine->id == RCS0 && INTEL_GEN(engine->i915) >= 9)
 		fw |= FORCEWAKE_RENDER;
 
 	intel_uncore_forcewake_get(uncore, fw);
 
 	intel_uncore_write_fw(uncore, reg, 0x1);
 
-	if (wait_for_atomic((intel_uncore_read_fw(uncore, reg) == 0), 50))
-		gvt_vgpu_err("timeout in invalidate ring (%d) tlb\n", ring_id);
+	if (wait_for_atomic(intel_uncore_read_fw(uncore, reg) == 0, 50))
+		gvt_vgpu_err("timeout in invalidate ring %s tlb\n",
+			     engine->name);
 	else
 		vgpu_vreg_t(vgpu, reg) = 0;
 
 	intel_uncore_forcewake_put(uncore, fw);
 
-	gvt_dbg_core("invalidate TLB for ring %d\n", ring_id);
+	gvt_dbg_core("invalidate TLB for ring %s\n", engine->name);
 }
 
 static void switch_mocs(struct intel_vgpu *pre, struct intel_vgpu *next,
-			int ring_id)
+			const struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv;
-	i915_reg_t offset, l3_offset;
-	u32 old_v, new_v;
-
 	u32 regs[] = {
 		[RCS0]  = 0xc800,
 		[VCS0]  = 0xc900,
@@ -403,36 +400,38 @@ static void switch_mocs(struct intel_vgpu *pre, struct intel_vgpu *next,
 		[BCS0]  = 0xcc00,
 		[VECS0] = 0xcb00,
 	};
+	struct intel_uncore *uncore = engine->uncore;
+	i915_reg_t offset, l3_offset;
+	u32 old_v, new_v;
 	int i;
 
-	dev_priv = pre ? pre->gvt->dev_priv : next->gvt->dev_priv;
-	if (WARN_ON(ring_id >= ARRAY_SIZE(regs)))
+	if (drm_WARN_ON(&engine->i915->drm, engine->id >= ARRAY_SIZE(regs)))
 		return;
 
-	if (ring_id == RCS0 && IS_GEN(dev_priv, 9))
+	if (engine->id == RCS0 && IS_GEN(engine->i915, 9))
 		return;
 
 	if (!pre && !gen9_render_mocs.initialized)
-		load_render_mocs(dev_priv);
+		load_render_mocs(engine);
 
-	offset.reg = regs[ring_id];
+	offset.reg = regs[engine->id];
 	for (i = 0; i < GEN9_MOCS_SIZE; i++) {
 		if (pre)
 			old_v = vgpu_vreg_t(pre, offset);
 		else
-			old_v = gen9_render_mocs.control_table[ring_id][i];
+			old_v = gen9_render_mocs.control_table[engine->id][i];
 		if (next)
 			new_v = vgpu_vreg_t(next, offset);
 		else
-			new_v = gen9_render_mocs.control_table[ring_id][i];
+			new_v = gen9_render_mocs.control_table[engine->id][i];
 
 		if (old_v != new_v)
-			I915_WRITE_FW(offset, new_v);
+			intel_uncore_write_fw(uncore, offset, new_v);
 
 		offset.reg += 4;
 	}
 
-	if (ring_id == RCS0) {
+	if (engine->id == RCS0) {
 		l3_offset.reg = 0xb020;
 		for (i = 0; i < GEN9_MOCS_SIZE / 2; i++) {
 			if (pre)
@@ -445,7 +444,7 @@ static void switch_mocs(struct intel_vgpu *pre, struct intel_vgpu *next,
 				new_v = gen9_render_mocs.l3cc_table[i];
 
 			if (old_v != new_v)
-				I915_WRITE_FW(l3_offset, new_v);
+				intel_uncore_write_fw(uncore, l3_offset, new_v);
 
 			l3_offset.reg += 4;
 		}
@@ -467,38 +466,40 @@ bool is_inhibit_context(struct intel_context *ce)
 /* Switch ring mmio values (context). */
 static void switch_mmio(struct intel_vgpu *pre,
 			struct intel_vgpu *next,
-			int ring_id)
+			const struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv;
+	struct intel_uncore *uncore = engine->uncore;
 	struct intel_vgpu_submission *s;
 	struct engine_mmio *mmio;
 	u32 old_v, new_v;
 
-	dev_priv = pre ? pre->gvt->dev_priv : next->gvt->dev_priv;
-	if (INTEL_GEN(dev_priv) >= 9)
-		switch_mocs(pre, next, ring_id);
+	if (INTEL_GEN(engine->i915) >= 9)
+		switch_mocs(pre, next, engine);
 
-	for (mmio = dev_priv->gvt->engine_mmio_list.mmio;
+	for (mmio = engine->i915->gvt->engine_mmio_list.mmio;
 	     i915_mmio_reg_valid(mmio->reg); mmio++) {
-		if (mmio->ring_id != ring_id)
+		if (mmio->id != engine->id)
 			continue;
 		/*
 		 * No need to do save or restore of the mmio which is in context
 		 * state image on gen9, it's initialized by lri command and
 		 * save or restore with context together.
 		 */
-		if (IS_GEN(dev_priv, 9) && mmio->in_context)
+		if (IS_GEN(engine->i915, 9) && mmio->in_context)
 			continue;
 
 		// save
 		if (pre) {
-			vgpu_vreg_t(pre, mmio->reg) = I915_READ_FW(mmio->reg);
+			vgpu_vreg_t(pre, mmio->reg) =
+				intel_uncore_read_fw(uncore, mmio->reg);
 			if (mmio->mask)
 				vgpu_vreg_t(pre, mmio->reg) &=
-						~(mmio->mask << 16);
+					~(mmio->mask << 16);
 			old_v = vgpu_vreg_t(pre, mmio->reg);
-		} else
-			old_v = mmio->value = I915_READ_FW(mmio->reg);
+		} else {
+			old_v = mmio->value =
+				intel_uncore_read_fw(uncore, mmio->reg);
+		}
 
 		// restore
 		if (next) {
@@ -509,12 +510,12 @@ static void switch_mmio(struct intel_vgpu *pre,
 			 * itself.
 			 */
 			if (mmio->in_context &&
-			    !is_inhibit_context(s->shadow[ring_id]))
+			    !is_inhibit_context(s->shadow[engine->id]))
 				continue;
 
 			if (mmio->mask)
 				new_v = vgpu_vreg_t(next, mmio->reg) |
-							(mmio->mask << 16);
+					(mmio->mask << 16);
 			else
 				new_v = vgpu_vreg_t(next, mmio->reg);
 		} else {
@@ -526,7 +527,7 @@ static void switch_mmio(struct intel_vgpu *pre,
 				new_v = mmio->value;
 		}
 
-		I915_WRITE_FW(mmio->reg, new_v);
+		intel_uncore_write_fw(uncore, mmio->reg, new_v);
 
 		trace_render_mmio(pre ? pre->id : 0,
 				  next ? next->id : 0,
@@ -536,39 +537,37 @@ static void switch_mmio(struct intel_vgpu *pre,
 	}
 
 	if (next)
-		handle_tlb_pending_event(next, ring_id);
+		handle_tlb_pending_event(next, engine);
 }
 
 /**
  * intel_gvt_switch_render_mmio - switch mmio context of specific engine
  * @pre: the last vGPU that own the engine
  * @next: the vGPU to switch to
- * @ring_id: specify the engine
+ * @engine: the engine
  *
  * If pre is null indicates that host own the engine. If next is null
  * indicates that we are switching to host workload.
  */
 void intel_gvt_switch_mmio(struct intel_vgpu *pre,
-			   struct intel_vgpu *next, int ring_id)
+			   struct intel_vgpu *next,
+			   const struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv;
-
-	if (WARN_ON(!pre && !next))
+	if (WARN(!pre && !next, "switch ring %s from host to HOST\n",
+		 engine->name))
 		return;
 
-	gvt_dbg_render("switch ring %d from %s to %s\n", ring_id,
+	gvt_dbg_render("switch ring %s from %s to %s\n", engine->name,
 		       pre ? "vGPU" : "host", next ? "vGPU" : "HOST");
 
-	dev_priv = pre ? pre->gvt->dev_priv : next->gvt->dev_priv;
-
 	/**
 	 * We are using raw mmio access wrapper to improve the
 	 * performace for batch mmio read/write, so we need
 	 * handle forcewake mannually.
 	 */
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-	switch_mmio(pre, next, ring_id);
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
+	intel_uncore_forcewake_get(engine->uncore, FORCEWAKE_ALL);
+	switch_mmio(pre, next, engine);
+	intel_uncore_forcewake_put(engine->uncore, FORCEWAKE_ALL);
 }
 
 /**
@@ -580,7 +579,7 @@ void intel_gvt_init_engine_mmio_context(struct intel_gvt *gvt)
 {
 	struct engine_mmio *mmio;
 
-	if (INTEL_GEN(gvt->dev_priv) >= 9) {
+	if (INTEL_GEN(gvt->gt->i915) >= 9) {
 		gvt->engine_mmio_list.mmio = gen9_engine_mmio_list;
 		gvt->engine_mmio_list.tlb_mmio_offset_list = gen8_tlb_mmio_offset_list;
 		gvt->engine_mmio_list.tlb_mmio_offset_list_cnt = ARRAY_SIZE(gen8_tlb_mmio_offset_list);
@@ -595,7 +594,7 @@ void intel_gvt_init_engine_mmio_context(struct intel_gvt *gvt)
 	for (mmio = gvt->engine_mmio_list.mmio;
 	     i915_mmio_reg_valid(mmio->reg); mmio++) {
 		if (mmio->in_context) {
-			gvt->engine_mmio_list.ctx_mmio_count[mmio->ring_id]++;
+			gvt->engine_mmio_list.ctx_mmio_count[mmio->id]++;
 			intel_gvt_mmio_set_in_ctx(gvt, mmio->reg.reg);
 		}
 	}
diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.h b/drivers/gpu/drm/i915/gvt/mmio_context.h
index f7eaa442403f..970704b18f23 100644
--- a/drivers/gpu/drm/i915/gvt/mmio_context.h
+++ b/drivers/gpu/drm/i915/gvt/mmio_context.h
@@ -37,7 +37,7 @@
 #define __GVT_RENDER_H__
 
 struct engine_mmio {
-	int ring_id;
+	enum intel_engine_id id;
 	i915_reg_t reg;
 	u32 mask;
 	bool in_context;
@@ -45,7 +45,8 @@ struct engine_mmio {
 };
 
 void intel_gvt_switch_mmio(struct intel_vgpu *pre,
-			   struct intel_vgpu *next, int ring_id);
+			   struct intel_vgpu *next,
+			   const struct intel_engine_cs *engine);
 
 void intel_gvt_init_engine_mmio_context(struct intel_gvt *gvt);
 
diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.c b/drivers/gpu/drm/i915/gvt/sched_policy.c
index 2369d4a9af94..036b74fe9298 100644
--- a/drivers/gpu/drm/i915/gvt/sched_policy.c
+++ b/drivers/gpu/drm/i915/gvt/sched_policy.c
@@ -39,8 +39,8 @@ static bool vgpu_has_pending_workload(struct intel_vgpu *vgpu)
 	enum intel_engine_id i;
 	struct intel_engine_cs *engine;
 
-	for_each_engine(engine, vgpu->gvt->dev_priv, i) {
-		if (!list_empty(workload_q_head(vgpu, i)))
+	for_each_engine(engine, vgpu->gvt->gt, i) {
+		if (!list_empty(workload_q_head(vgpu, engine)))
 			return true;
 	}
 
@@ -152,8 +152,8 @@ static void try_to_schedule_next_vgpu(struct intel_gvt *gvt)
 	scheduler->need_reschedule = true;
 
 	/* still have uncompleted workload? */
-	for_each_engine(engine, gvt->dev_priv, i) {
-		if (scheduler->current_workload[i])
+	for_each_engine(engine, gvt->gt, i) {
+		if (scheduler->current_workload[engine->id])
 			return;
 	}
 
@@ -169,8 +169,8 @@ static void try_to_schedule_next_vgpu(struct intel_gvt *gvt)
 	scheduler->need_reschedule = false;
 
 	/* wake up workload dispatch thread */
-	for_each_engine(engine, gvt->dev_priv, i)
-		wake_up(&scheduler->waitq[i]);
+	for_each_engine(engine, gvt->gt, i)
+		wake_up(&scheduler->waitq[engine->id]);
 }
 
 static struct intel_vgpu *find_busy_vgpu(struct gvt_sched_data *sched_data)
@@ -444,9 +444,10 @@ void intel_vgpu_stop_schedule(struct intel_vgpu *vgpu)
 {
 	struct intel_gvt_workload_scheduler *scheduler =
 		&vgpu->gvt->scheduler;
-	int ring_id;
 	struct vgpu_sched_data *vgpu_data = vgpu->sched_data;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
 
 	if (!vgpu_data->active)
 		return;
@@ -467,10 +468,10 @@ void intel_vgpu_stop_schedule(struct intel_vgpu *vgpu)
 
 	intel_runtime_pm_get(&dev_priv->runtime_pm);
 	spin_lock_bh(&scheduler->mmio_context_lock);
-	for (ring_id = 0; ring_id < I915_NUM_ENGINES; ring_id++) {
-		if (scheduler->engine_owner[ring_id] == vgpu) {
-			intel_gvt_switch_mmio(vgpu, NULL, ring_id);
-			scheduler->engine_owner[ring_id] = NULL;
+	for_each_engine(engine, vgpu->gvt->gt, id) {
+		if (scheduler->engine_owner[engine->id] == vgpu) {
+			intel_gvt_switch_mmio(vgpu, NULL, engine);
+			scheduler->engine_owner[engine->id] = NULL;
 		}
 	}
 	spin_unlock_bh(&scheduler->mmio_context_lock);
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index e8c0885df978..1c95bf8cbed0 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -84,7 +84,7 @@ static void update_shadow_pdps(struct intel_vgpu_workload *workload)
 static void sr_oa_regs(struct intel_vgpu_workload *workload,
 		u32 *reg_state, bool save)
 {
-	struct drm_i915_private *dev_priv = workload->vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = workload->vgpu->gvt->gt->i915;
 	u32 ctx_oactxctrl = dev_priv->perf.ctx_oactxctrl_offset;
 	u32 ctx_flexeu0 = dev_priv->perf.ctx_flexeu0_offset;
 	int i = 0;
@@ -98,7 +98,7 @@ static void sr_oa_regs(struct intel_vgpu_workload *workload,
 		i915_mmio_reg_offset(EU_PERF_CNTL6),
 	};
 
-	if (workload->ring_id != RCS0)
+	if (workload->engine->id != RCS0)
 		return;
 
 	if (save) {
@@ -128,7 +128,6 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct intel_gvt *gvt = vgpu->gvt;
-	int ring_id = workload->ring_id;
 	struct drm_i915_gem_object *ctx_obj =
 		workload->req->context->state->obj;
 	struct execlist_ring_context *shadow_ring_context;
@@ -154,7 +153,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	COPY_REG_MASKED(ctx_ctrl);
 	COPY_REG(ctx_timestamp);
 
-	if (ring_id == RCS0) {
+	if (workload->engine->id == RCS0) {
 		COPY_REG(bb_per_ctx_ptr);
 		COPY_REG(rcs_indirect_ctx);
 		COPY_REG(rcs_indirect_ctx_offset);
@@ -175,14 +174,14 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	if (IS_RESTORE_INHIBIT(shadow_ring_context->ctx_ctrl.val))
 		return 0;
 
-	gvt_dbg_sched("ring id %d workload lrca %x", ring_id,
-			workload->ctx_desc.lrca);
-
-	context_page_num = gvt->dev_priv->engine[ring_id]->context_size;
+	gvt_dbg_sched("ring %s workload lrca %x",
+		      workload->engine->name,
+		      workload->ctx_desc.lrca);
 
+	context_page_num = workload->engine->context_size;
 	context_page_num = context_page_num >> PAGE_SHIFT;
 
-	if (IS_BROADWELL(gvt->dev_priv) && ring_id == RCS0)
+	if (IS_BROADWELL(gvt->gt->i915) && workload->engine->id == RCS0)
 		context_page_num = 19;
 
 	i = 2;
@@ -210,38 +209,43 @@ static inline bool is_gvt_request(struct i915_request *rq)
 	return intel_context_force_single_submission(rq->context);
 }
 
-static void save_ring_hw_state(struct intel_vgpu *vgpu, int ring_id)
+static void save_ring_hw_state(struct intel_vgpu *vgpu,
+			       const struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
-	u32 ring_base = dev_priv->engine[ring_id]->mmio_base;
+	struct intel_uncore *uncore = engine->uncore;
 	i915_reg_t reg;
 
-	reg = RING_INSTDONE(ring_base);
-	vgpu_vreg(vgpu, i915_mmio_reg_offset(reg)) = I915_READ_FW(reg);
-	reg = RING_ACTHD(ring_base);
-	vgpu_vreg(vgpu, i915_mmio_reg_offset(reg)) = I915_READ_FW(reg);
-	reg = RING_ACTHD_UDW(ring_base);
-	vgpu_vreg(vgpu, i915_mmio_reg_offset(reg)) = I915_READ_FW(reg);
+	reg = RING_INSTDONE(engine->mmio_base);
+	vgpu_vreg(vgpu, i915_mmio_reg_offset(reg)) =
+		intel_uncore_read(uncore, reg);
+
+	reg = RING_ACTHD(engine->mmio_base);
+	vgpu_vreg(vgpu, i915_mmio_reg_offset(reg)) =
+		intel_uncore_read(uncore, reg);
+
+	reg = RING_ACTHD_UDW(engine->mmio_base);
+	vgpu_vreg(vgpu, i915_mmio_reg_offset(reg)) =
+		intel_uncore_read(uncore, reg);
 }
 
 static int shadow_context_status_change(struct notifier_block *nb,
 		unsigned long action, void *data)
 {
-	struct i915_request *req = data;
+	struct i915_request *rq = data;
 	struct intel_gvt *gvt = container_of(nb, struct intel_gvt,
-				shadow_ctx_notifier_block[req->engine->id]);
+				shadow_ctx_notifier_block[rq->engine->id]);
 	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
-	enum intel_engine_id ring_id = req->engine->id;
+	enum intel_engine_id ring_id = rq->engine->id;
 	struct intel_vgpu_workload *workload;
 	unsigned long flags;
 
-	if (!is_gvt_request(req)) {
+	if (!is_gvt_request(rq)) {
 		spin_lock_irqsave(&scheduler->mmio_context_lock, flags);
 		if (action == INTEL_CONTEXT_SCHEDULE_IN &&
 		    scheduler->engine_owner[ring_id]) {
 			/* Switch ring from vGPU to host. */
 			intel_gvt_switch_mmio(scheduler->engine_owner[ring_id],
-					      NULL, ring_id);
+					      NULL, rq->engine);
 			scheduler->engine_owner[ring_id] = NULL;
 		}
 		spin_unlock_irqrestore(&scheduler->mmio_context_lock, flags);
@@ -259,7 +263,7 @@ static int shadow_context_status_change(struct notifier_block *nb,
 		if (workload->vgpu != scheduler->engine_owner[ring_id]) {
 			/* Switch ring from host to vGPU or vGPU to vGPU. */
 			intel_gvt_switch_mmio(scheduler->engine_owner[ring_id],
-					      workload->vgpu, ring_id);
+					      workload->vgpu, rq->engine);
 			scheduler->engine_owner[ring_id] = workload->vgpu;
 		} else
 			gvt_dbg_sched("skip ring %d mmio switch for vgpu%d\n",
@@ -268,11 +272,11 @@ static int shadow_context_status_change(struct notifier_block *nb,
 		atomic_set(&workload->shadow_ctx_active, 1);
 		break;
 	case INTEL_CONTEXT_SCHEDULE_OUT:
-		save_ring_hw_state(workload->vgpu, ring_id);
+		save_ring_hw_state(workload->vgpu, rq->engine);
 		atomic_set(&workload->shadow_ctx_active, 0);
 		break;
 	case INTEL_CONTEXT_SCHEDULE_PREEMPTED:
-		save_ring_hw_state(workload->vgpu, ring_id);
+		save_ring_hw_state(workload->vgpu, rq->engine);
 		break;
 	default:
 		WARN_ON(1);
@@ -391,7 +395,7 @@ intel_gvt_workload_req_alloc(struct intel_vgpu_workload *workload)
 	if (workload->req)
 		return 0;
 
-	rq = i915_request_create(s->shadow[workload->ring_id]);
+	rq = i915_request_create(s->shadow[workload->engine->id]);
 	if (IS_ERR(rq)) {
 		gvt_vgpu_err("fail to allocate gem request\n");
 		return PTR_ERR(rq);
@@ -420,15 +424,16 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	if (workload->shadow)
 		return 0;
 
-	if (!test_and_set_bit(workload->ring_id, s->shadow_ctx_desc_updated))
-		shadow_context_descriptor_update(s->shadow[workload->ring_id],
+	if (!test_and_set_bit(workload->engine->id, s->shadow_ctx_desc_updated))
+		shadow_context_descriptor_update(s->shadow[workload->engine->id],
 						 workload);
 
 	ret = intel_gvt_scan_and_shadow_ringbuffer(workload);
 	if (ret)
 		return ret;
 
-	if (workload->ring_id == RCS0 && workload->wa_ctx.indirect_ctx.size) {
+	if (workload->engine->id == RCS0 &&
+	    workload->wa_ctx.indirect_ctx.size) {
 		ret = intel_gvt_scan_and_shadow_wa_ctx(&workload->wa_ctx);
 		if (ret)
 			goto err_shadow;
@@ -436,6 +441,7 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 
 	workload->shadow = true;
 	return 0;
+
 err_shadow:
 	release_shadow_wa_ctx(&workload->wa_ctx);
 	return ret;
@@ -567,12 +573,8 @@ static int prepare_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 
 static void update_vreg_in_ctx(struct intel_vgpu_workload *workload)
 {
-	struct intel_vgpu *vgpu = workload->vgpu;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
-	u32 ring_base;
-
-	ring_base = dev_priv->engine[workload->ring_id]->mmio_base;
-	vgpu_vreg_t(vgpu, RING_START(ring_base)) = workload->rb_start;
+	vgpu_vreg_t(workload->vgpu, RING_START(workload->engine->mmio_base)) =
+		workload->rb_start;
 }
 
 static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload)
@@ -608,7 +610,6 @@ static int prepare_workload(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	int ring = workload->ring_id;
 	int ret = 0;
 
 	ret = intel_vgpu_pin_mm(workload->shadow_mm);
@@ -625,7 +626,7 @@ static int prepare_workload(struct intel_vgpu_workload *workload)
 
 	update_shadow_pdps(workload);
 
-	set_context_ppgtt_from_shadow(workload, s->shadow[ring]);
+	set_context_ppgtt_from_shadow(workload, s->shadow[workload->engine->id]);
 
 	ret = intel_vgpu_sync_oos_pages(workload->vgpu);
 	if (ret) {
@@ -677,11 +678,10 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct i915_request *rq;
-	int ring_id = workload->ring_id;
 	int ret;
 
-	gvt_dbg_sched("ring id %d prepare to dispatch workload %p\n",
-		ring_id, workload);
+	gvt_dbg_sched("ring id %s prepare to dispatch workload %p\n",
+		      workload->engine->name, workload);
 
 	mutex_lock(&vgpu->vgpu_lock);
 
@@ -710,8 +710,8 @@ out:
 	}
 
 	if (!IS_ERR_OR_NULL(workload->req)) {
-		gvt_dbg_sched("ring id %d submit workload to i915 %p\n",
-				ring_id, workload->req);
+		gvt_dbg_sched("ring id %s submit workload to i915 %p\n",
+			      workload->engine->name, workload->req);
 		i915_request_add(workload->req);
 		workload->dispatched = true;
 	}
@@ -722,8 +722,8 @@ err_req:
 	return ret;
 }
 
-static struct intel_vgpu_workload *pick_next_workload(
-		struct intel_gvt *gvt, int ring_id)
+static struct intel_vgpu_workload *
+pick_next_workload(struct intel_gvt *gvt, struct intel_engine_cs *engine)
 {
 	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
 	struct intel_vgpu_workload *workload = NULL;
@@ -735,27 +735,27 @@ static struct intel_vgpu_workload *pick_next_workload(
 	 * bail out
 	 */
 	if (!scheduler->current_vgpu) {
-		gvt_dbg_sched("ring id %d stop - no current vgpu\n", ring_id);
+		gvt_dbg_sched("ring %s stop - no current vgpu\n", engine->name);
 		goto out;
 	}
 
 	if (scheduler->need_reschedule) {
-		gvt_dbg_sched("ring id %d stop - will reschedule\n", ring_id);
+		gvt_dbg_sched("ring %s stop - will reschedule\n", engine->name);
 		goto out;
 	}
 
 	if (!scheduler->current_vgpu->active ||
-	    list_empty(workload_q_head(scheduler->current_vgpu, ring_id)))
+	    list_empty(workload_q_head(scheduler->current_vgpu, engine)))
 		goto out;
 
 	/*
 	 * still have current workload, maybe the workload disptacher
 	 * fail to submit it for some reason, resubmit it.
 	 */
-	if (scheduler->current_workload[ring_id]) {
-		workload = scheduler->current_workload[ring_id];
-		gvt_dbg_sched("ring id %d still have current workload %p\n",
-				ring_id, workload);
+	if (scheduler->current_workload[engine->id]) {
+		workload = scheduler->current_workload[engine->id];
+		gvt_dbg_sched("ring %s still have current workload %p\n",
+			      engine->name, workload);
 		goto out;
 	}
 
@@ -765,13 +765,14 @@ static struct intel_vgpu_workload *pick_next_workload(
 	 * will wait the current workload is finished when trying to
 	 * schedule out a vgpu.
 	 */
-	scheduler->current_workload[ring_id] = container_of(
-			workload_q_head(scheduler->current_vgpu, ring_id)->next,
-			struct intel_vgpu_workload, list);
+	scheduler->current_workload[engine->id] =
+		list_first_entry(workload_q_head(scheduler->current_vgpu,
+						 engine),
+				 struct intel_vgpu_workload, list);
 
-	workload = scheduler->current_workload[ring_id];
+	workload = scheduler->current_workload[engine->id];
 
-	gvt_dbg_sched("ring id %d pick new workload %p\n", ring_id, workload);
+	gvt_dbg_sched("ring %s pick new workload %p\n", engine->name, workload);
 
 	atomic_inc(&workload->vgpu->submission.running_workload_num);
 out:
@@ -783,14 +784,12 @@ 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->context->state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *src;
 	unsigned long context_gpa, context_page_num;
 	int i;
-	struct drm_i915_private *dev_priv = gvt->dev_priv;
 	u32 ring_base;
 	u32 head, tail;
 	u16 wrap_count;
@@ -811,14 +810,14 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 
 	head = (wrap_count << RB_HEAD_WRAP_CNT_OFF) | tail;
 
-	ring_base = dev_priv->engine[workload->ring_id]->mmio_base;
+	ring_base = rq->engine->mmio_base;
 	vgpu_vreg_t(vgpu, RING_TAIL(ring_base)) = tail;
 	vgpu_vreg_t(vgpu, RING_HEAD(ring_base)) = head;
 
 	context_page_num = rq->engine->context_size;
 	context_page_num = context_page_num >> PAGE_SHIFT;
 
-	if (IS_BROADWELL(gvt->dev_priv) && rq->engine->id == RCS0)
+	if (IS_BROADWELL(rq->i915) && rq->engine->id == RCS0)
 		context_page_num = 19;
 
 	i = 2;
@@ -869,7 +868,7 @@ void intel_vgpu_clean_workloads(struct intel_vgpu *vgpu,
 				intel_engine_mask_t engine_mask)
 {
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	struct drm_i915_private *dev_priv = vgpu->gvt->gt->i915;
 	struct intel_engine_cs *engine;
 	struct intel_vgpu_workload *pos, *n;
 	intel_engine_mask_t tmp;
@@ -966,54 +965,47 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 	mutex_unlock(&vgpu->vgpu_lock);
 }
 
-struct workload_thread_param {
-	struct intel_gvt *gvt;
-	int ring_id;
-};
-
-static int workload_thread(void *priv)
+static int workload_thread(void *arg)
 {
-	struct workload_thread_param *p = (struct workload_thread_param *)priv;
-	struct intel_gvt *gvt = p->gvt;
-	int ring_id = p->ring_id;
+	struct intel_engine_cs *engine = arg;
+	const bool need_force_wake = INTEL_GEN(engine->i915) >= 9;
+	struct intel_gvt *gvt = engine->i915->gvt;
 	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
 	struct intel_vgpu_workload *workload = NULL;
 	struct intel_vgpu *vgpu = NULL;
 	int ret;
-	bool need_force_wake = (INTEL_GEN(gvt->dev_priv) >= 9);
 	DEFINE_WAIT_FUNC(wait, woken_wake_function);
-	struct intel_runtime_pm *rpm = &gvt->dev_priv->runtime_pm;
-
-	kfree(p);
 
-	gvt_dbg_core("workload thread for ring %d started\n", ring_id);
+	gvt_dbg_core("workload thread for ring %s started\n", engine->name);
 
 	while (!kthread_should_stop()) {
-		add_wait_queue(&scheduler->waitq[ring_id], &wait);
+		intel_wakeref_t wakeref;
+
+		add_wait_queue(&scheduler->waitq[engine->id], &wait);
 		do {
-			workload = pick_next_workload(gvt, ring_id);
+			workload = pick_next_workload(gvt, engine);
 			if (workload)
 				break;
 			wait_woken(&wait, TASK_INTERRUPTIBLE,
 				   MAX_SCHEDULE_TIMEOUT);
 		} while (!kthread_should_stop());
-		remove_wait_queue(&scheduler->waitq[ring_id], &wait);
+		remove_wait_queue(&scheduler->waitq[engine->id], &wait);
 
 		if (!workload)
 			break;
 
-		gvt_dbg_sched("ring id %d next workload %p vgpu %d\n",
-				workload->ring_id, workload,
-				workload->vgpu->id);
+		gvt_dbg_sched("ring %s next workload %p vgpu %d\n",
+			      engine->name, workload,
+			      workload->vgpu->id);
 
-		intel_runtime_pm_get(rpm);
+		wakeref = intel_runtime_pm_get(engine->uncore->rpm);
 
-		gvt_dbg_sched("ring id %d will dispatch workload %p\n",
-				workload->ring_id, workload);
+		gvt_dbg_sched("ring %s will dispatch workload %p\n",
+			      engine->name, workload);
 
 		if (need_force_wake)
-			intel_uncore_forcewake_get(&gvt->dev_priv->uncore,
-					FORCEWAKE_ALL);
+			intel_uncore_forcewake_get(engine->uncore,
+						   FORCEWAKE_ALL);
 		/*
 		 * Update the vReg of the vGPU which submitted this
 		 * workload. The vGPU may use these registers for checking
@@ -1030,21 +1022,21 @@ static int workload_thread(void *priv)
 			goto complete;
 		}
 
-		gvt_dbg_sched("ring id %d wait workload %p\n",
-				workload->ring_id, workload);
+		gvt_dbg_sched("ring %s wait workload %p\n",
+			      engine->name, workload);
 		i915_request_wait(workload->req, 0, MAX_SCHEDULE_TIMEOUT);
 
 complete:
 		gvt_dbg_sched("will complete workload %p, status: %d\n",
-				workload, workload->status);
+			      workload, workload->status);
 
-		complete_current_workload(gvt, ring_id);
+		complete_current_workload(gvt, engine->id);
 
 		if (need_force_wake)
-			intel_uncore_forcewake_put(&gvt->dev_priv->uncore,
-					FORCEWAKE_ALL);
+			intel_uncore_forcewake_put(engine->uncore,
+						   FORCEWAKE_ALL);
 
-		intel_runtime_pm_put_unchecked(rpm);
+		intel_runtime_pm_put(engine->uncore->rpm, wakeref);
 		if (ret && (vgpu_is_vm_unhealthy(ret)))
 			enter_failsafe_mode(vgpu, GVT_FAILSAFE_GUEST_ERR);
 	}
@@ -1073,7 +1065,7 @@ void intel_gvt_clean_workload_scheduler(struct intel_gvt *gvt)
 
 	gvt_dbg_core("clean workload scheduler\n");
 
-	for_each_engine(engine, gvt->dev_priv, i) {
+	for_each_engine(engine, gvt->gt, i) {
 		atomic_notifier_chain_unregister(
 					&engine->context_status_notifier,
 					&gvt->shadow_ctx_notifier_block[i]);
@@ -1084,7 +1076,6 @@ void intel_gvt_clean_workload_scheduler(struct intel_gvt *gvt)
 int intel_gvt_init_workload_scheduler(struct intel_gvt *gvt)
 {
 	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
-	struct workload_thread_param *param = NULL;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id i;
 	int ret;
@@ -1093,20 +1084,11 @@ int intel_gvt_init_workload_scheduler(struct intel_gvt *gvt)
 
 	init_waitqueue_head(&scheduler->workload_complete_wq);
 
-	for_each_engine(engine, gvt->dev_priv, i) {
+	for_each_engine(engine, gvt->gt, i) {
 		init_waitqueue_head(&scheduler->waitq[i]);
 
-		param = kzalloc(sizeof(*param), GFP_KERNEL);
-		if (!param) {
-			ret = -ENOMEM;
-			goto err;
-		}
-
-		param->gvt = gvt;
-		param->ring_id = i;
-
-		scheduler->thread[i] = kthread_run(workload_thread, param,
-			"gvt workload %d", i);
+		scheduler->thread[i] = kthread_run(workload_thread, engine,
+						   "gvt:%s", engine->name);
 		if (IS_ERR(scheduler->thread[i])) {
 			gvt_err("fail to create workload thread\n");
 			ret = PTR_ERR(scheduler->thread[i]);
@@ -1118,11 +1100,11 @@ int intel_gvt_init_workload_scheduler(struct intel_gvt *gvt)
 		atomic_notifier_chain_register(&engine->context_status_notifier,
 					&gvt->shadow_ctx_notifier_block[i]);
 	}
+
 	return 0;
+
 err:
 	intel_gvt_clean_workload_scheduler(gvt);
-	kfree(param);
-	param = NULL;
 	return ret;
 }
 
@@ -1160,7 +1142,7 @@ void intel_vgpu_clean_submission(struct intel_vgpu *vgpu)
 	intel_vgpu_select_submission_ops(vgpu, ALL_ENGINES, 0);
 
 	i915_context_ppgtt_root_restore(s, i915_vm_to_ppgtt(s->shadow[0]->vm));
-	for_each_engine(engine, vgpu->gvt->dev_priv, id)
+	for_each_engine(engine, vgpu->gvt->gt, id)
 		intel_context_unpin(s->shadow[id]);
 
 	kmem_cache_destroy(s->workloads);
@@ -1217,7 +1199,7 @@ i915_context_ppgtt_root_save(struct intel_vgpu_submission *s,
  */
 int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 {
-	struct drm_i915_private *i915 = vgpu->gvt->dev_priv;
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	struct intel_engine_cs *engine;
 	struct i915_ppgtt *ppgtt;
@@ -1230,7 +1212,7 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 
 	i915_context_ppgtt_root_save(s, ppgtt);
 
-	for_each_engine(engine, i915, i) {
+	for_each_engine(engine, vgpu->gvt->gt, i) {
 		struct intel_context *ce;
 
 		INIT_LIST_HEAD(&s->workload_q_head[i]);
@@ -1283,7 +1265,7 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 
 out_shadow_ctx:
 	i915_context_ppgtt_root_restore(s, ppgtt);
-	for_each_engine(engine, i915, i) {
+	for_each_engine(engine, vgpu->gvt->gt, i) {
 		if (IS_ERR(s->shadow[i]))
 			break;
 
@@ -1310,6 +1292,7 @@ int intel_vgpu_select_submission_ops(struct intel_vgpu *vgpu,
 				     intel_engine_mask_t engine_mask,
 				     unsigned int interface)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	const struct intel_vgpu_submission_ops *ops[] = {
 		[INTEL_VGPU_EXECLIST_SUBMISSION] =
@@ -1317,10 +1300,11 @@ int intel_vgpu_select_submission_ops(struct intel_vgpu *vgpu,
 	};
 	int ret;
 
-	if (WARN_ON(interface >= ARRAY_SIZE(ops)))
+	if (drm_WARN_ON(&i915->drm, interface >= ARRAY_SIZE(ops)))
 		return -EINVAL;
 
-	if (WARN_ON(interface == 0 && engine_mask != ALL_ENGINES))
+	if (drm_WARN_ON(&i915->drm,
+			interface == 0 && engine_mask != ALL_ENGINES))
 		return -EINVAL;
 
 	if (s->active)
@@ -1442,7 +1426,7 @@ static int prepare_mm(struct intel_vgpu_workload *workload)
 /**
  * intel_vgpu_create_workload - create a vGPU workload
  * @vgpu: a vGPU
- * @ring_id: ring index
+ * @engine: the engine
  * @desc: a guest context descriptor
  *
  * This function is called when creating a vGPU workload.
@@ -1453,14 +1437,14 @@ static int prepare_mm(struct intel_vgpu_workload *workload)
  *
  */
 struct intel_vgpu_workload *
-intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
+intel_vgpu_create_workload(struct intel_vgpu *vgpu,
+			   const struct intel_engine_cs *engine,
 			   struct execlist_ctx_descriptor_format *desc)
 {
 	struct intel_vgpu_submission *s = &vgpu->submission;
-	struct list_head *q = workload_q_head(vgpu, ring_id);
+	struct list_head *q = workload_q_head(vgpu, engine);
 	struct intel_vgpu_workload *last_workload = NULL;
 	struct intel_vgpu_workload *workload = NULL;
-	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	u64 ring_context_gpa;
 	u32 head, tail, start, ctl, ctx_ctl, per_ctx, indirect_ctx;
 	u32 guest_head;
@@ -1487,10 +1471,10 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 	list_for_each_entry_reverse(last_workload, q, list) {
 
 		if (same_context(&last_workload->ctx_desc, desc)) {
-			gvt_dbg_el("ring id %d cur workload == last\n",
-					ring_id);
+			gvt_dbg_el("ring %s cur workload == last\n",
+				   engine->name);
 			gvt_dbg_el("ctx head %x real head %lx\n", head,
-					last_workload->rb_tail);
+				   last_workload->rb_tail);
 			/*
 			 * cannot use guest context head pointer here,
 			 * as it might not be updated at this time
@@ -1500,7 +1484,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 		}
 	}
 
-	gvt_dbg_el("ring id %d begin a new workload\n", ring_id);
+	gvt_dbg_el("ring %s begin a new workload\n", engine->name);
 
 	/* record some ring buffer register values for scan and shadow */
 	intel_gvt_hypervisor_read_gpa(vgpu, ring_context_gpa +
@@ -1520,7 +1504,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 	if (IS_ERR(workload))
 		return workload;
 
-	workload->ring_id = ring_id;
+	workload->engine = engine;
 	workload->ctx_desc = *desc;
 	workload->ring_context_gpa = ring_context_gpa;
 	workload->rb_head = head;
@@ -1529,7 +1513,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 	workload->rb_start = start;
 	workload->rb_ctl = ctl;
 
-	if (ring_id == RCS0) {
+	if (engine->id == RCS0) {
 		intel_gvt_hypervisor_read_gpa(vgpu, ring_context_gpa +
 			RING_CTX_OFF(bb_per_ctx_ptr.val), &per_ctx, 4);
 		intel_gvt_hypervisor_read_gpa(vgpu, ring_context_gpa +
@@ -1567,8 +1551,8 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 		}
 	}
 
-	gvt_dbg_el("workload %p ring id %d head %x tail %x start %x ctl %x\n",
-			workload, ring_id, head, tail, start, ctl);
+	gvt_dbg_el("workload %p ring %s head %x tail %x start %x ctl %x\n",
+		   workload, engine->name, head, tail, start, ctl);
 
 	ret = prepare_mm(workload);
 	if (ret) {
@@ -1579,10 +1563,11 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 	/* Only scan and shadow the first workload in the queue
 	 * as there is only one pre-allocated buf-obj for shadow.
 	 */
-	if (list_empty(workload_q_head(vgpu, ring_id))) {
-		intel_runtime_pm_get(&dev_priv->runtime_pm);
-		ret = intel_gvt_scan_and_shadow_workload(workload);
-		intel_runtime_pm_put_unchecked(&dev_priv->runtime_pm);
+	if (list_empty(q)) {
+		intel_wakeref_t wakeref;
+
+		with_intel_runtime_pm(engine->gt->uncore->rpm, wakeref)
+			ret = intel_gvt_scan_and_shadow_workload(workload);
 	}
 
 	if (ret) {
@@ -1602,7 +1587,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 void intel_vgpu_queue_workload(struct intel_vgpu_workload *workload)
 {
 	list_add_tail(&workload->list,
-		workload_q_head(workload->vgpu, workload->ring_id));
+		      workload_q_head(workload->vgpu, workload->engine));
 	intel_gvt_kick_schedule(workload->vgpu->gvt);
-	wake_up(&workload->vgpu->gvt->scheduler.waitq[workload->ring_id]);
+	wake_up(&workload->vgpu->gvt->scheduler.waitq[workload->engine->id]);
 }
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.h b/drivers/gpu/drm/i915/gvt/scheduler.h
index c50d14a9ce85..bf7fc0ca4cb1 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.h
+++ b/drivers/gpu/drm/i915/gvt/scheduler.h
@@ -79,7 +79,7 @@ struct intel_shadow_wa_ctx {
 
 struct intel_vgpu_workload {
 	struct intel_vgpu *vgpu;
-	int ring_id;
+	const struct intel_engine_cs *engine;
 	struct i915_request *req;
 	/* if this workload has been dispatched to i915? */
 	bool dispatched;
@@ -129,8 +129,8 @@ struct intel_vgpu_shadow_bb {
 	bool ppgtt;
 };
 
-#define workload_q_head(vgpu, ring_id) \
-	(&(vgpu->submission.workload_q_head[ring_id]))
+#define workload_q_head(vgpu, e) \
+	(&(vgpu)->submission.workload_q_head[(e)->id])
 
 void intel_vgpu_queue_workload(struct intel_vgpu_workload *workload);
 
@@ -155,7 +155,8 @@ extern const struct intel_vgpu_submission_ops
 intel_vgpu_execlist_submission_ops;
 
 struct intel_vgpu_workload *
-intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
+intel_vgpu_create_workload(struct intel_vgpu *vgpu,
+			   const struct intel_engine_cs *engine,
 			   struct execlist_ctx_descriptor_format *desc);
 
 void intel_vgpu_destroy_workload(struct intel_vgpu_workload *workload);
diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c
index 487af6ea9972..78f14f04d2ea 100644
--- a/drivers/gpu/drm/i915/gvt/vgpu.c
+++ b/drivers/gpu/drm/i915/gvt/vgpu.c
@@ -37,6 +37,7 @@
 
 void populate_pvinfo_page(struct intel_vgpu *vgpu)
 {
+	struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
 	/* setup the ballooning information */
 	vgpu_vreg64_t(vgpu, vgtif_reg(magic)) = VGT_MAGIC;
 	vgpu_vreg_t(vgpu, vgtif_reg(version_major)) = 1;
@@ -69,7 +70,7 @@ void populate_pvinfo_page(struct intel_vgpu *vgpu)
 		vgpu_hidden_gmadr_base(vgpu), vgpu_hidden_sz(vgpu));
 	gvt_dbg_core("fence size %d\n", vgpu_fence_sz(vgpu));
 
-	WARN_ON(sizeof(struct vgt_if) != VGT_PVINFO_SIZE);
+	drm_WARN_ON(&i915->drm, sizeof(struct vgt_if) != VGT_PVINFO_SIZE);
 }
 
 #define VGPU_MAX_WEIGHT 16
@@ -148,12 +149,12 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt)
 		gvt->types[i].avail_instance = min(low_avail / vgpu_types[i].low_mm,
 						   high_avail / vgpu_types[i].high_mm);
 
-		if (IS_GEN(gvt->dev_priv, 8))
+		if (IS_GEN(gvt->gt->i915, 8))
 			sprintf(gvt->types[i].name, "GVTg_V4_%s",
-						vgpu_types[i].name);
-		else if (IS_GEN(gvt->dev_priv, 9))
+				vgpu_types[i].name);
+		else if (IS_GEN(gvt->gt->i915, 9))
 			sprintf(gvt->types[i].name, "GVTg_V5_%s",
-						vgpu_types[i].name);
+				vgpu_types[i].name);
 
 		gvt_dbg_core("type[%d]: %s avail %u low %u high %u fence %u weight %u res %s\n",
 			     i, gvt->types[i].name,
@@ -271,10 +272,11 @@ void intel_gvt_release_vgpu(struct intel_vgpu *vgpu)
 void intel_gvt_destroy_vgpu(struct intel_vgpu *vgpu)
 {
 	struct intel_gvt *gvt = vgpu->gvt;
+	struct drm_i915_private *i915 = gvt->gt->i915;
 
 	mutex_lock(&vgpu->vgpu_lock);
 
-	WARN(vgpu->active, "vGPU is still active!\n");
+	drm_WARN(&i915->drm, vgpu->active, "vGPU is still active!\n");
 
 	intel_gvt_debugfs_remove_vgpu(vgpu);
 	intel_vgpu_clean_sched_policy(vgpu);
@@ -426,9 +428,7 @@ static struct intel_vgpu *__intel_gvt_create_vgpu(struct intel_gvt *gvt,
 	if (ret)
 		goto out_clean_sched_policy;
 
-	/*TODO: add more platforms support */
-	if (IS_SKYLAKE(gvt->dev_priv) || IS_KABYLAKE(gvt->dev_priv))
-		ret = intel_gvt_hypervisor_set_edid(vgpu, PORT_D);
+	ret = intel_gvt_hypervisor_set_edid(vgpu, PORT_D);
 	if (ret)
 		goto out_clean_sched_policy;
 
diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c
index 992b00fc5745..c4048628188a 100644
--- a/drivers/gpu/drm/i915/i915_active.c
+++ b/drivers/gpu/drm/i915/i915_active.c
@@ -7,6 +7,7 @@
 #include <linux/debugobjects.h>
 
 #include "gt/intel_context.h"
+#include "gt/intel_engine_heartbeat.h"
 #include "gt/intel_engine_pm.h"
 #include "gt/intel_ring.h"
 
@@ -452,6 +453,9 @@ static void enable_signaling(struct i915_active_fence *active)
 {
 	struct dma_fence *fence;
 
+	if (unlikely(is_barrier(active)))
+		return;
+
 	fence = i915_active_fence_get(active);
 	if (!fence)
 		return;
@@ -460,26 +464,49 @@ static void enable_signaling(struct i915_active_fence *active)
 	dma_fence_put(fence);
 }
 
-int i915_active_wait(struct i915_active *ref)
+static int flush_barrier(struct active_node *it)
 {
-	struct active_node *it, *n;
-	int err = 0;
+	struct intel_engine_cs *engine;
 
-	might_sleep();
+	if (likely(!is_barrier(&it->base)))
+		return 0;
 
-	if (!i915_active_acquire_if_busy(ref))
+	engine = __barrier_to_engine(it);
+	smp_rmb(); /* serialise with add_active_barriers */
+	if (!is_barrier(&it->base))
 		return 0;
 
-	/* Flush lazy signals */
+	return intel_engine_flush_barriers(engine);
+}
+
+static int flush_lazy_signals(struct i915_active *ref)
+{
+	struct active_node *it, *n;
+	int err = 0;
+
 	enable_signaling(&ref->excl);
 	rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
-		if (is_barrier(&it->base)) /* unconnected idle barrier */
-			continue;
+		err = flush_barrier(it); /* unconnected idle barrier? */
+		if (err)
+			break;
 
 		enable_signaling(&it->base);
 	}
-	/* Any fence added after the wait begins will not be auto-signaled */
 
+	return err;
+}
+
+int i915_active_wait(struct i915_active *ref)
+{
+	int err;
+
+	might_sleep();
+
+	if (!i915_active_acquire_if_busy(ref))
+		return 0;
+
+	/* Any fence added after the wait begins will not be auto-signaled */
+	err = flush_lazy_signals(ref);
 	i915_active_release(ref);
 	if (err)
 		return err;
@@ -491,25 +518,81 @@ int i915_active_wait(struct i915_active *ref)
 	return 0;
 }
 
-int i915_request_await_active(struct i915_request *rq, struct i915_active *ref)
+static int __await_active(struct i915_active_fence *active,
+			  int (*fn)(void *arg, struct dma_fence *fence),
+			  void *arg)
+{
+	struct dma_fence *fence;
+
+	if (is_barrier(active)) /* XXX flush the barrier? */
+		return 0;
+
+	fence = i915_active_fence_get(active);
+	if (fence) {
+		int err;
+
+		err = fn(arg, fence);
+		dma_fence_put(fence);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
+}
+
+static int await_active(struct i915_active *ref,
+			unsigned int flags,
+			int (*fn)(void *arg, struct dma_fence *fence),
+			void *arg)
 {
 	int err = 0;
 
+	/* We must always wait for the exclusive fence! */
 	if (rcu_access_pointer(ref->excl.fence)) {
-		struct dma_fence *fence;
-
-		rcu_read_lock();
-		fence = dma_fence_get_rcu_safe(&ref->excl.fence);
-		rcu_read_unlock();
-		if (fence) {
-			err = i915_request_await_dma_fence(rq, fence);
-			dma_fence_put(fence);
+		err = __await_active(&ref->excl, fn, arg);
+		if (err)
+			return err;
+	}
+
+	if (flags & I915_ACTIVE_AWAIT_ALL && i915_active_acquire_if_busy(ref)) {
+		struct active_node *it, *n;
+
+		rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
+			err = __await_active(&it->base, fn, arg);
+			if (err)
+				break;
 		}
+		i915_active_release(ref);
+		if (err)
+			return err;
 	}
 
-	/* In the future we may choose to await on all fences */
+	return 0;
+}
+
+static int rq_await_fence(void *arg, struct dma_fence *fence)
+{
+	return i915_request_await_dma_fence(arg, fence);
+}
 
-	return err;
+int i915_request_await_active(struct i915_request *rq,
+			      struct i915_active *ref,
+			      unsigned int flags)
+{
+	return await_active(ref, flags, rq_await_fence, rq);
+}
+
+static int sw_await_fence(void *arg, struct dma_fence *fence)
+{
+	return i915_sw_fence_await_dma_fence(arg, fence, 0,
+					     GFP_NOWAIT | __GFP_NOWARN);
+}
+
+int i915_sw_fence_await_active(struct i915_sw_fence *fence,
+			       struct i915_active *ref,
+			       unsigned int flags)
+{
+	return await_active(ref, flags, sw_await_fence, fence);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
diff --git a/drivers/gpu/drm/i915/i915_active.h b/drivers/gpu/drm/i915/i915_active.h
index 973ff0447c6c..b3282ae7913c 100644
--- a/drivers/gpu/drm/i915/i915_active.h
+++ b/drivers/gpu/drm/i915/i915_active.h
@@ -183,7 +183,13 @@ static inline bool i915_active_has_exclusive(struct i915_active *ref)
 
 int i915_active_wait(struct i915_active *ref);
 
-int i915_request_await_active(struct i915_request *rq, struct i915_active *ref);
+int i915_sw_fence_await_active(struct i915_sw_fence *fence,
+			       struct i915_active *ref,
+			       unsigned int flags);
+int i915_request_await_active(struct i915_request *rq,
+			      struct i915_active *ref,
+			      unsigned int flags);
+#define I915_ACTIVE_AWAIT_ALL BIT(0)
 
 int i915_active_acquire(struct i915_active *ref);
 bool i915_active_acquire_if_busy(struct i915_active *ref);
diff --git a/drivers/gpu/drm/i915/i915_buddy.c b/drivers/gpu/drm/i915/i915_buddy.c
index 66883af64ca1..20babbdb297d 100644
--- a/drivers/gpu/drm/i915/i915_buddy.c
+++ b/drivers/gpu/drm/i915/i915_buddy.c
@@ -312,7 +312,8 @@ i915_buddy_alloc(struct i915_buddy_mm *mm, unsigned int order)
 	return block;
 
 out_free:
-	__i915_buddy_free(mm, block);
+	if (i != order)
+		__i915_buddy_free(mm, block);
 	return ERR_PTR(err);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 8f2525e4ce0f..6ca797128aa1 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -996,220 +996,6 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 	return ret;
 }
 
-static int ilk_drpc_info(struct seq_file *m)
-{
-	struct drm_i915_private *i915 = node_to_i915(m->private);
-	struct intel_uncore *uncore = &i915->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 i915_forcewake_domains(struct seq_file *m, void *data)
-{
-	struct drm_i915_private *i915 = node_to_i915(m->private);
-	struct intel_uncore *uncore = &i915->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;
-}
-
-static void print_rc6_res(struct seq_file *m,
-			  const char *title,
-			  const i915_reg_t reg)
-{
-	struct drm_i915_private *i915 = node_to_i915(m->private);
-	intel_wakeref_t wakeref;
-
-	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
-		seq_printf(m, "%s %u (%llu us)\n", title,
-			   intel_uncore_read(&i915->uncore, reg),
-			   intel_rc6_residency_us(&i915->gt.rc6, reg));
-}
-
-static int vlv_drpc_info(struct seq_file *m)
-{
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	u32 rcctl1, pw_status;
-
-	pw_status = I915_READ(VLV_GTLC_PW_STATUS);
-	rcctl1 = I915_READ(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 i915_forcewake_domains(m, NULL);
-}
-
-static int gen6_drpc_info(struct seq_file *m)
-{
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	u32 gt_core_status, rcctl1, rc6vids = 0;
-	u32 gen9_powergate_enable = 0, gen9_powergate_status = 0;
-
-	gt_core_status = I915_READ_FW(GEN6_GT_CORE_STATUS);
-	trace_i915_reg_rw(false, GEN6_GT_CORE_STATUS, gt_core_status, 4, true);
-
-	rcctl1 = I915_READ(GEN6_RC_CONTROL);
-	if (INTEL_GEN(dev_priv) >= 9) {
-		gen9_powergate_enable = I915_READ(GEN9_PG_ENABLE);
-		gen9_powergate_status = I915_READ(GEN9_PWRGT_DOMAIN_STATUS);
-	}
-
-	if (INTEL_GEN(dev_priv) <= 7)
-		sandybridge_pcode_read(dev_priv, 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(dev_priv) >= 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(dev_priv) >= 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(dev_priv) <= 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 i915_forcewake_domains(m, NULL);
-}
-
-static int i915_drpc_info(struct seq_file *m, void *unused)
-{
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	intel_wakeref_t wakeref;
-	int err = -ENODEV;
-
-	with_intel_runtime_pm(&dev_priv->runtime_pm, wakeref) {
-		if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-			err = vlv_drpc_info(m);
-		else if (INTEL_GEN(dev_priv) >= 6)
-			err = gen6_drpc_info(m);
-		else
-			err = ilk_drpc_info(m);
-	}
-
-	return err;
-}
-
 static int i915_ring_freq_table(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -2360,10 +2146,8 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_guc_stage_pool", i915_guc_stage_pool, 0},
 	{"i915_huc_load_status", i915_huc_load_status_info, 0},
 	{"i915_frequency_info", i915_frequency_info, 0},
-	{"i915_drpc_info", i915_drpc_info, 0},
 	{"i915_ring_freq_table", i915_ring_freq_table, 0},
 	{"i915_context_status", i915_context_status, 0},
-	{"i915_forcewake_domains", i915_forcewake_domains, 0},
 	{"i915_swizzle_info", i915_swizzle_info, 0},
 	{"i915_llc", i915_llc, 0},
 	{"i915_runtime_pm_status", i915_runtime_pm_status, 0},
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index dba5fe1391e8..82d9df15b22b 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -44,7 +44,6 @@
 #include <drm/drm_ioctl.h>
 #include <drm/drm_irq.h>
 #include <drm/drm_probe_helper.h>
-#include <drm/i915_drm.h>
 
 #include "display/intel_acpi.h"
 #include "display/intel_audio.h"
@@ -71,6 +70,7 @@
 
 #include "i915_debugfs.h"
 #include "i915_drv.h"
+#include "i915_ioc32.h"
 #include "i915_irq.h"
 #include "i915_memcpy.h"
 #include "i915_perf.h"
@@ -80,6 +80,8 @@
 #include "i915_sysfs.h"
 #include "i915_trace.h"
 #include "i915_vgpu.h"
+#include "intel_dram.h"
+#include "intel_gvt.h"
 #include "intel_memory_region.h"
 #include "intel_pm.h"
 #include "vlv_suspend.h"
@@ -238,8 +240,14 @@ static int i915_driver_modeset_probe_noirq(struct drm_i915_private *i915)
 
 	intel_csr_ucode_init(i915);
 
+	ret = intel_modeset_init_noirq(i915);
+	if (ret)
+		goto cleanup_vga_client;
+
 	return 0;
 
+cleanup_vga_client:
+	intel_vga_unregister(i915);
 out:
 	return ret;
 }
@@ -381,6 +389,7 @@ static void intel_detect_preproduction_hw(struct drm_i915_private *dev_priv)
 	pre |= IS_SKL_REVID(dev_priv, 0, SKL_REVID_F0);
 	pre |= IS_BXT_REVID(dev_priv, 0, BXT_REVID_B_LAST);
 	pre |= IS_KBL_REVID(dev_priv, 0, KBL_REVID_A0);
+	pre |= IS_GLK_REVID(dev_priv, 0, GLK_REVID_A2);
 
 	if (pre) {
 		drm_err(&dev_priv->drm, "This is a pre-production stepping. "
@@ -459,7 +468,6 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
 	intel_init_display_hooks(dev_priv);
 	intel_init_clock_gating_hooks(dev_priv);
 	intel_init_audio_hooks(dev_priv);
-	intel_display_crc_init(dev_priv);
 
 	intel_detect_preproduction_hw(dev_priv);
 
@@ -558,494 +566,6 @@ static void intel_sanitize_options(struct drm_i915_private *dev_priv)
 	intel_gvt_sanitize_options(dev_priv);
 }
 
-#define DRAM_TYPE_STR(type) [INTEL_DRAM_ ## type] = #type
-
-static const char *intel_dram_type_str(enum intel_dram_type type)
-{
-	static const char * const str[] = {
-		DRAM_TYPE_STR(UNKNOWN),
-		DRAM_TYPE_STR(DDR3),
-		DRAM_TYPE_STR(DDR4),
-		DRAM_TYPE_STR(LPDDR3),
-		DRAM_TYPE_STR(LPDDR4),
-	};
-
-	if (type >= ARRAY_SIZE(str))
-		type = INTEL_DRAM_UNKNOWN;
-
-	return str[type];
-}
-
-#undef DRAM_TYPE_STR
-
-static int intel_dimm_num_devices(const struct dram_dimm_info *dimm)
-{
-	return dimm->ranks * 64 / (dimm->width ?: 1);
-}
-
-/* Returns total GB for the whole DIMM */
-static int skl_get_dimm_size(u16 val)
-{
-	return val & SKL_DRAM_SIZE_MASK;
-}
-
-static int skl_get_dimm_width(u16 val)
-{
-	if (skl_get_dimm_size(val) == 0)
-		return 0;
-
-	switch (val & SKL_DRAM_WIDTH_MASK) {
-	case SKL_DRAM_WIDTH_X8:
-	case SKL_DRAM_WIDTH_X16:
-	case SKL_DRAM_WIDTH_X32:
-		val = (val & SKL_DRAM_WIDTH_MASK) >> SKL_DRAM_WIDTH_SHIFT;
-		return 8 << val;
-	default:
-		MISSING_CASE(val);
-		return 0;
-	}
-}
-
-static int skl_get_dimm_ranks(u16 val)
-{
-	if (skl_get_dimm_size(val) == 0)
-		return 0;
-
-	val = (val & SKL_DRAM_RANK_MASK) >> SKL_DRAM_RANK_SHIFT;
-
-	return val + 1;
-}
-
-/* Returns total GB for the whole DIMM */
-static int cnl_get_dimm_size(u16 val)
-{
-	return (val & CNL_DRAM_SIZE_MASK) / 2;
-}
-
-static int cnl_get_dimm_width(u16 val)
-{
-	if (cnl_get_dimm_size(val) == 0)
-		return 0;
-
-	switch (val & CNL_DRAM_WIDTH_MASK) {
-	case CNL_DRAM_WIDTH_X8:
-	case CNL_DRAM_WIDTH_X16:
-	case CNL_DRAM_WIDTH_X32:
-		val = (val & CNL_DRAM_WIDTH_MASK) >> CNL_DRAM_WIDTH_SHIFT;
-		return 8 << val;
-	default:
-		MISSING_CASE(val);
-		return 0;
-	}
-}
-
-static int cnl_get_dimm_ranks(u16 val)
-{
-	if (cnl_get_dimm_size(val) == 0)
-		return 0;
-
-	val = (val & CNL_DRAM_RANK_MASK) >> CNL_DRAM_RANK_SHIFT;
-
-	return val + 1;
-}
-
-static bool
-skl_is_16gb_dimm(const struct dram_dimm_info *dimm)
-{
-	/* Convert total GB to Gb per DRAM device */
-	return 8 * dimm->size / (intel_dimm_num_devices(dimm) ?: 1) == 16;
-}
-
-static void
-skl_dram_get_dimm_info(struct drm_i915_private *dev_priv,
-		       struct dram_dimm_info *dimm,
-		       int channel, char dimm_name, u16 val)
-{
-	if (INTEL_GEN(dev_priv) >= 10) {
-		dimm->size = cnl_get_dimm_size(val);
-		dimm->width = cnl_get_dimm_width(val);
-		dimm->ranks = cnl_get_dimm_ranks(val);
-	} else {
-		dimm->size = skl_get_dimm_size(val);
-		dimm->width = skl_get_dimm_width(val);
-		dimm->ranks = skl_get_dimm_ranks(val);
-	}
-
-	drm_dbg_kms(&dev_priv->drm,
-		    "CH%u DIMM %c size: %u GB, width: X%u, ranks: %u, 16Gb DIMMs: %s\n",
-		    channel, dimm_name, dimm->size, dimm->width, dimm->ranks,
-		    yesno(skl_is_16gb_dimm(dimm)));
-}
-
-static int
-skl_dram_get_channel_info(struct drm_i915_private *dev_priv,
-			  struct dram_channel_info *ch,
-			  int channel, u32 val)
-{
-	skl_dram_get_dimm_info(dev_priv, &ch->dimm_l,
-			       channel, 'L', val & 0xffff);
-	skl_dram_get_dimm_info(dev_priv, &ch->dimm_s,
-			       channel, 'S', val >> 16);
-
-	if (ch->dimm_l.size == 0 && ch->dimm_s.size == 0) {
-		drm_dbg_kms(&dev_priv->drm, "CH%u not populated\n", channel);
-		return -EINVAL;
-	}
-
-	if (ch->dimm_l.ranks == 2 || ch->dimm_s.ranks == 2)
-		ch->ranks = 2;
-	else if (ch->dimm_l.ranks == 1 && ch->dimm_s.ranks == 1)
-		ch->ranks = 2;
-	else
-		ch->ranks = 1;
-
-	ch->is_16gb_dimm =
-		skl_is_16gb_dimm(&ch->dimm_l) ||
-		skl_is_16gb_dimm(&ch->dimm_s);
-
-	drm_dbg_kms(&dev_priv->drm, "CH%u ranks: %u, 16Gb DIMMs: %s\n",
-		    channel, ch->ranks, yesno(ch->is_16gb_dimm));
-
-	return 0;
-}
-
-static bool
-intel_is_dram_symmetric(const struct dram_channel_info *ch0,
-			const struct dram_channel_info *ch1)
-{
-	return !memcmp(ch0, ch1, sizeof(*ch0)) &&
-		(ch0->dimm_s.size == 0 ||
-		 !memcmp(&ch0->dimm_l, &ch0->dimm_s, sizeof(ch0->dimm_l)));
-}
-
-static int
-skl_dram_get_channels_info(struct drm_i915_private *dev_priv)
-{
-	struct dram_info *dram_info = &dev_priv->dram_info;
-	struct dram_channel_info ch0 = {}, ch1 = {};
-	u32 val;
-	int ret;
-
-	val = I915_READ(SKL_MAD_DIMM_CH0_0_0_0_MCHBAR_MCMAIN);
-	ret = skl_dram_get_channel_info(dev_priv, &ch0, 0, val);
-	if (ret == 0)
-		dram_info->num_channels++;
-
-	val = I915_READ(SKL_MAD_DIMM_CH1_0_0_0_MCHBAR_MCMAIN);
-	ret = skl_dram_get_channel_info(dev_priv, &ch1, 1, val);
-	if (ret == 0)
-		dram_info->num_channels++;
-
-	if (dram_info->num_channels == 0) {
-		drm_info(&dev_priv->drm,
-			 "Number of memory channels is zero\n");
-		return -EINVAL;
-	}
-
-	/*
-	 * If any of the channel is single rank channel, worst case output
-	 * will be same as if single rank memory, so consider single rank
-	 * memory.
-	 */
-	if (ch0.ranks == 1 || ch1.ranks == 1)
-		dram_info->ranks = 1;
-	else
-		dram_info->ranks = max(ch0.ranks, ch1.ranks);
-
-	if (dram_info->ranks == 0) {
-		drm_info(&dev_priv->drm,
-			 "couldn't get memory rank information\n");
-		return -EINVAL;
-	}
-
-	dram_info->is_16gb_dimm = ch0.is_16gb_dimm || ch1.is_16gb_dimm;
-
-	dram_info->symmetric_memory = intel_is_dram_symmetric(&ch0, &ch1);
-
-	drm_dbg_kms(&dev_priv->drm, "Memory configuration is symmetric? %s\n",
-		    yesno(dram_info->symmetric_memory));
-	return 0;
-}
-
-static enum intel_dram_type
-skl_get_dram_type(struct drm_i915_private *dev_priv)
-{
-	u32 val;
-
-	val = I915_READ(SKL_MAD_INTER_CHANNEL_0_0_0_MCHBAR_MCMAIN);
-
-	switch (val & SKL_DRAM_DDR_TYPE_MASK) {
-	case SKL_DRAM_DDR_TYPE_DDR3:
-		return INTEL_DRAM_DDR3;
-	case SKL_DRAM_DDR_TYPE_DDR4:
-		return INTEL_DRAM_DDR4;
-	case SKL_DRAM_DDR_TYPE_LPDDR3:
-		return INTEL_DRAM_LPDDR3;
-	case SKL_DRAM_DDR_TYPE_LPDDR4:
-		return INTEL_DRAM_LPDDR4;
-	default:
-		MISSING_CASE(val);
-		return INTEL_DRAM_UNKNOWN;
-	}
-}
-
-static int
-skl_get_dram_info(struct drm_i915_private *dev_priv)
-{
-	struct dram_info *dram_info = &dev_priv->dram_info;
-	u32 mem_freq_khz, val;
-	int ret;
-
-	dram_info->type = skl_get_dram_type(dev_priv);
-	drm_dbg_kms(&dev_priv->drm, "DRAM type: %s\n",
-		    intel_dram_type_str(dram_info->type));
-
-	ret = skl_dram_get_channels_info(dev_priv);
-	if (ret)
-		return ret;
-
-	val = I915_READ(SKL_MC_BIOS_DATA_0_0_0_MCHBAR_PCU);
-	mem_freq_khz = DIV_ROUND_UP((val & SKL_REQ_DATA_MASK) *
-				    SKL_MEMORY_FREQ_MULTIPLIER_HZ, 1000);
-
-	dram_info->bandwidth_kbps = dram_info->num_channels *
-							mem_freq_khz * 8;
-
-	if (dram_info->bandwidth_kbps == 0) {
-		drm_info(&dev_priv->drm,
-			 "Couldn't get system memory bandwidth\n");
-		return -EINVAL;
-	}
-
-	dram_info->valid = true;
-	return 0;
-}
-
-/* Returns Gb per DRAM device */
-static int bxt_get_dimm_size(u32 val)
-{
-	switch (val & BXT_DRAM_SIZE_MASK) {
-	case BXT_DRAM_SIZE_4GBIT:
-		return 4;
-	case BXT_DRAM_SIZE_6GBIT:
-		return 6;
-	case BXT_DRAM_SIZE_8GBIT:
-		return 8;
-	case BXT_DRAM_SIZE_12GBIT:
-		return 12;
-	case BXT_DRAM_SIZE_16GBIT:
-		return 16;
-	default:
-		MISSING_CASE(val);
-		return 0;
-	}
-}
-
-static int bxt_get_dimm_width(u32 val)
-{
-	if (!bxt_get_dimm_size(val))
-		return 0;
-
-	val = (val & BXT_DRAM_WIDTH_MASK) >> BXT_DRAM_WIDTH_SHIFT;
-
-	return 8 << val;
-}
-
-static int bxt_get_dimm_ranks(u32 val)
-{
-	if (!bxt_get_dimm_size(val))
-		return 0;
-
-	switch (val & BXT_DRAM_RANK_MASK) {
-	case BXT_DRAM_RANK_SINGLE:
-		return 1;
-	case BXT_DRAM_RANK_DUAL:
-		return 2;
-	default:
-		MISSING_CASE(val);
-		return 0;
-	}
-}
-
-static enum intel_dram_type bxt_get_dimm_type(u32 val)
-{
-	if (!bxt_get_dimm_size(val))
-		return INTEL_DRAM_UNKNOWN;
-
-	switch (val & BXT_DRAM_TYPE_MASK) {
-	case BXT_DRAM_TYPE_DDR3:
-		return INTEL_DRAM_DDR3;
-	case BXT_DRAM_TYPE_LPDDR3:
-		return INTEL_DRAM_LPDDR3;
-	case BXT_DRAM_TYPE_DDR4:
-		return INTEL_DRAM_DDR4;
-	case BXT_DRAM_TYPE_LPDDR4:
-		return INTEL_DRAM_LPDDR4;
-	default:
-		MISSING_CASE(val);
-		return INTEL_DRAM_UNKNOWN;
-	}
-}
-
-static void bxt_get_dimm_info(struct dram_dimm_info *dimm,
-			      u32 val)
-{
-	dimm->width = bxt_get_dimm_width(val);
-	dimm->ranks = bxt_get_dimm_ranks(val);
-
-	/*
-	 * Size in register is Gb per DRAM device. Convert to total
-	 * GB to match the way we report this for non-LP platforms.
-	 */
-	dimm->size = bxt_get_dimm_size(val) * intel_dimm_num_devices(dimm) / 8;
-}
-
-static int
-bxt_get_dram_info(struct drm_i915_private *dev_priv)
-{
-	struct dram_info *dram_info = &dev_priv->dram_info;
-	u32 dram_channels;
-	u32 mem_freq_khz, val;
-	u8 num_active_channels;
-	int i;
-
-	val = I915_READ(BXT_P_CR_MC_BIOS_REQ_0_0_0);
-	mem_freq_khz = DIV_ROUND_UP((val & BXT_REQ_DATA_MASK) *
-				    BXT_MEMORY_FREQ_MULTIPLIER_HZ, 1000);
-
-	dram_channels = val & BXT_DRAM_CHANNEL_ACTIVE_MASK;
-	num_active_channels = hweight32(dram_channels);
-
-	/* Each active bit represents 4-byte channel */
-	dram_info->bandwidth_kbps = (mem_freq_khz * num_active_channels * 4);
-
-	if (dram_info->bandwidth_kbps == 0) {
-		drm_info(&dev_priv->drm,
-			 "Couldn't get system memory bandwidth\n");
-		return -EINVAL;
-	}
-
-	/*
-	 * Now read each DUNIT8/9/10/11 to check the rank of each dimms.
-	 */
-	for (i = BXT_D_CR_DRP0_DUNIT_START; i <= BXT_D_CR_DRP0_DUNIT_END; i++) {
-		struct dram_dimm_info dimm;
-		enum intel_dram_type type;
-
-		val = I915_READ(BXT_D_CR_DRP0_DUNIT(i));
-		if (val == 0xFFFFFFFF)
-			continue;
-
-		dram_info->num_channels++;
-
-		bxt_get_dimm_info(&dimm, val);
-		type = bxt_get_dimm_type(val);
-
-		drm_WARN_ON(&dev_priv->drm, type != INTEL_DRAM_UNKNOWN &&
-			    dram_info->type != INTEL_DRAM_UNKNOWN &&
-			    dram_info->type != type);
-
-		drm_dbg_kms(&dev_priv->drm,
-			    "CH%u DIMM size: %u GB, width: X%u, ranks: %u, type: %s\n",
-			    i - BXT_D_CR_DRP0_DUNIT_START,
-			    dimm.size, dimm.width, dimm.ranks,
-			    intel_dram_type_str(type));
-
-		/*
-		 * If any of the channel is single rank channel,
-		 * worst case output will be same as if single rank
-		 * memory, so consider single rank memory.
-		 */
-		if (dram_info->ranks == 0)
-			dram_info->ranks = dimm.ranks;
-		else if (dimm.ranks == 1)
-			dram_info->ranks = 1;
-
-		if (type != INTEL_DRAM_UNKNOWN)
-			dram_info->type = type;
-	}
-
-	if (dram_info->type == INTEL_DRAM_UNKNOWN ||
-	    dram_info->ranks == 0) {
-		drm_info(&dev_priv->drm, "couldn't get memory information\n");
-		return -EINVAL;
-	}
-
-	dram_info->valid = true;
-	return 0;
-}
-
-static void
-intel_get_dram_info(struct drm_i915_private *dev_priv)
-{
-	struct dram_info *dram_info = &dev_priv->dram_info;
-	int ret;
-
-	/*
-	 * Assume 16Gb DIMMs are present until proven otherwise.
-	 * This is only used for the level 0 watermark latency
-	 * w/a which does not apply to bxt/glk.
-	 */
-	dram_info->is_16gb_dimm = !IS_GEN9_LP(dev_priv);
-
-	if (INTEL_GEN(dev_priv) < 9 || !HAS_DISPLAY(dev_priv))
-		return;
-
-	if (IS_GEN9_LP(dev_priv))
-		ret = bxt_get_dram_info(dev_priv);
-	else
-		ret = skl_get_dram_info(dev_priv);
-	if (ret)
-		return;
-
-	drm_dbg_kms(&dev_priv->drm, "DRAM bandwidth: %u kBps, channels: %u\n",
-		    dram_info->bandwidth_kbps,
-		    dram_info->num_channels);
-
-	drm_dbg_kms(&dev_priv->drm, "DRAM ranks: %u, 16Gb DIMMs: %s\n",
-		    dram_info->ranks, yesno(dram_info->is_16gb_dimm));
-}
-
-static u32 gen9_edram_size_mb(struct drm_i915_private *dev_priv, u32 cap)
-{
-	static const u8 ways[8] = { 4, 8, 12, 16, 16, 16, 16, 16 };
-	static const u8 sets[4] = { 1, 1, 2, 2 };
-
-	return EDRAM_NUM_BANKS(cap) *
-		ways[EDRAM_WAYS_IDX(cap)] *
-		sets[EDRAM_SETS_IDX(cap)];
-}
-
-static void edram_detect(struct drm_i915_private *dev_priv)
-{
-	u32 edram_cap = 0;
-
-	if (!(IS_HASWELL(dev_priv) ||
-	      IS_BROADWELL(dev_priv) ||
-	      INTEL_GEN(dev_priv) >= 9))
-		return;
-
-	edram_cap = __raw_uncore_read32(&dev_priv->uncore, HSW_EDRAM_CAP);
-
-	/* NB: We can't write IDICR yet because we don't have gt funcs set up */
-
-	if (!(edram_cap & EDRAM_ENABLED))
-		return;
-
-	/*
-	 * The needed capability bits for size calculation are not there with
-	 * pre gen9 so return 128MB always.
-	 */
-	if (INTEL_GEN(dev_priv) < 9)
-		dev_priv->edram_size_mb = 128;
-	else
-		dev_priv->edram_size_mb =
-			gen9_edram_size_mb(dev_priv, edram_cap);
-
-	dev_info(dev_priv->drm.dev,
-		 "Found %uMB of eDRAM\n", dev_priv->edram_size_mb);
-}
-
 /**
  * i915_driver_hw_probe - setup state requiring device access
  * @dev_priv: device private
@@ -1089,7 +609,7 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 	intel_sanitize_options(dev_priv);
 
 	/* needs to be done before ggtt probe */
-	edram_detect(dev_priv);
+	intel_dram_edram_detect(dev_priv);
 
 	i915_perf_init(dev_priv);
 
@@ -1191,7 +711,7 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 	 * Fill the dram structure to get the system raw bandwidth and
 	 * dram info. This will be used for memory latency calculation.
 	 */
-	intel_get_dram_info(dev_priv);
+	intel_dram_detect(dev_priv);
 
 	intel_bw_init_hw(dev_priv);
 
@@ -1240,12 +760,7 @@ static void i915_driver_register(struct drm_i915_private *dev_priv)
 	i915_gem_driver_register(dev_priv);
 	i915_pmu_register(dev_priv);
 
-	/*
-	 * Notify a valid surface after modesetting,
-	 * when running inside a VM.
-	 */
-	if (intel_vgpu_active(dev_priv))
-		I915_WRITE(vgtif_reg(display_ready), VGT_DRV_DISPLAY_READY);
+	intel_vgpu_register(dev_priv);
 
 	/* Reveal our presence to userspace */
 	if (drm_dev_register(dev, 0) == 0) {
@@ -1375,8 +890,6 @@ i915_driver_create(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return ERR_PTR(err);
 	}
 
-	i915->drm.dev_private = i915;
-
 	i915->drm.pdev = pdev;
 	pci_set_drvdata(pdev, i915);
 
@@ -1454,7 +967,7 @@ int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	disable_rpm_wakeref_asserts(&i915->runtime_pm);
 
-	i915_detect_vgpu(i915);
+	intel_vgpu_detect(i915);
 
 	ret = i915_driver_mmio_probe(i915);
 	if (ret < 0)
@@ -2245,7 +1758,7 @@ static const struct file_operations i915_driver_fops = {
 	.mmap = i915_gem_mmap,
 	.poll = drm_poll,
 	.read = drm_read,
-	.compat_ioctl = i915_compat_ioctl,
+	.compat_ioctl = i915_ioc32_compat_ioctl,
 	.llseek = noop_llseek,
 };
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index ea13fc0b409b..1f5b9a584f71 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -59,7 +59,6 @@
 #include <drm/drm_connector.h>
 #include <drm/i915_mei_hdcp_interface.h>
 
-#include "i915_fixed.h"
 #include "i915_params.h"
 #include "i915_reg.h"
 #include "i915_utils.h"
@@ -105,18 +104,23 @@
 
 #include "intel_region_lmem.h"
 
-#include "intel_gvt.h"
-
 /* General customization:
  */
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20200225"
-#define DRIVER_TIMESTAMP	1582656081
+#define DRIVER_DATE		"20200313"
+#define DRIVER_TIMESTAMP	1584144591
 
 struct drm_i915_gem_object;
 
+/*
+ * The code assumes that the hpd_pins below have consecutive values and
+ * starting with HPD_PORT_A, the HPD pin associated with any port can be
+ * retrieved by adding the corresponding port (or phy) enum value to
+ * HPD_PORT_A in most cases. For example:
+ * HPD_PORT_C = HPD_PORT_A + PHY_C - PHY_A
+ */
 enum hpd_pin {
 	HPD_NONE = 0,
 	HPD_TV = HPD_NONE,     /* TV is known to be unreliable */
@@ -732,14 +736,6 @@ enum intel_ddb_partitioning {
 	INTEL_DDB_PART_5_6, /* IVB+ */
 };
 
-struct intel_wm_level {
-	bool enable;
-	u32 pri_val;
-	u32 spr_val;
-	u32 cur_val;
-	u32 fbc_val;
-};
-
 struct ilk_wm_values {
 	u32 wm_pipe[3];
 	u32 wm_lp[3];
@@ -798,56 +794,6 @@ static inline bool skl_ddb_entry_equal(const struct skl_ddb_entry *e1,
 	return false;
 }
 
-struct skl_wm_level {
-	u16 min_ddb_alloc;
-	u16 plane_res_b;
-	u8 plane_res_l;
-	bool plane_en;
-	bool ignore_lines;
-};
-
-/* Stores plane specific WM parameters */
-struct skl_wm_params {
-	bool x_tiled, y_tiled;
-	bool rc_surface;
-	bool is_planar;
-	u32 width;
-	u8 cpp;
-	u32 plane_pixel_rate;
-	u32 y_min_scanlines;
-	u32 plane_bytes_per_line;
-	uint_fixed_16_16_t plane_blocks_per_line;
-	uint_fixed_16_16_t y_tile_minimum;
-	u32 linetime_us;
-	u32 dbuf_block_size;
-};
-
-enum intel_pipe_crc_source {
-	INTEL_PIPE_CRC_SOURCE_NONE,
-	INTEL_PIPE_CRC_SOURCE_PLANE1,
-	INTEL_PIPE_CRC_SOURCE_PLANE2,
-	INTEL_PIPE_CRC_SOURCE_PLANE3,
-	INTEL_PIPE_CRC_SOURCE_PLANE4,
-	INTEL_PIPE_CRC_SOURCE_PLANE5,
-	INTEL_PIPE_CRC_SOURCE_PLANE6,
-	INTEL_PIPE_CRC_SOURCE_PLANE7,
-	INTEL_PIPE_CRC_SOURCE_PIPE,
-	/* TV/DP on pre-gen5/vlv can't use the pipe source. */
-	INTEL_PIPE_CRC_SOURCE_TV,
-	INTEL_PIPE_CRC_SOURCE_DP_B,
-	INTEL_PIPE_CRC_SOURCE_DP_C,
-	INTEL_PIPE_CRC_SOURCE_DP_D,
-	INTEL_PIPE_CRC_SOURCE_AUTO,
-	INTEL_PIPE_CRC_SOURCE_MAX,
-};
-
-#define INTEL_PIPE_CRC_ENTRIES_NR	128
-struct intel_pipe_crc {
-	spinlock_t lock;
-	int skipped;
-	enum intel_pipe_crc_source source;
-};
-
 struct i915_frontbuffer_tracking {
 	spinlock_t lock;
 
@@ -865,13 +811,6 @@ struct i915_virtual_gpu {
 	u32 caps;
 };
 
-/* used in computing the new watermarks state */
-struct intel_wm_config {
-	unsigned int num_pipes_active;
-	bool sprites_enabled;
-	bool sprites_scaled;
-};
-
 struct intel_cdclk_config {
 	unsigned int cdclk, vco, ref, bypass;
 	u8 voltage_level;
@@ -1043,21 +982,24 @@ struct drm_i915_private {
 	struct intel_crtc *plane_to_crtc_mapping[I915_MAX_PIPES];
 	struct intel_crtc *pipe_to_crtc_mapping[I915_MAX_PIPES];
 
-#ifdef CONFIG_DEBUG_FS
-	struct intel_pipe_crc pipe_crc[I915_MAX_PIPES];
-#endif
+	/**
+	 * dpll and cdclk state is protected by connection_mutex
+	 * dpll.lock serializes intel_{prepare,enable,disable}_shared_dpll.
+	 * Must be global rather than per dpll, because on some platforms plls
+	 * share registers.
+	 */
+	struct {
+		struct mutex lock;
 
-	/* dpll and cdclk state is protected by connection_mutex */
-	int num_shared_dpll;
-	struct intel_shared_dpll shared_dplls[I915_NUM_PLLS];
-	const struct intel_dpll_mgr *dpll_mgr;
+		int num_shared_dpll;
+		struct intel_shared_dpll shared_dplls[I915_NUM_PLLS];
+		const struct intel_dpll_mgr *mgr;
 
-	/*
-	 * dpll_lock serializes intel_{prepare,enable,disable}_shared_dpll.
-	 * Must be global rather than per dpll, because on some platforms
-	 * plls share registers.
-	 */
-	struct mutex dpll_lock;
+		struct {
+			int nssc;
+			int ssc;
+		} ref_clks;
+	} dpll;
 
 	struct list_head global_obj_list;
 
@@ -1078,8 +1020,6 @@ struct drm_i915_private {
 		struct work_struct free_work;
 	} atomic_helper;
 
-	u16 orig_clock;
-
 	bool mchbar_need_disable;
 
 	struct intel_l3_parity l3_parity;
@@ -1274,16 +1214,6 @@ struct drm_i915_private {
 	 */
 };
 
-struct dram_dimm_info {
-	u8 size, width, ranks;
-};
-
-struct dram_channel_info {
-	struct dram_dimm_info dimm_l, dimm_s;
-	u8 ranks;
-	bool is_16gb_dimm;
-};
-
 static inline struct drm_i915_private *to_i915(const struct drm_device *dev)
 {
 	return container_of(dev, struct drm_i915_private, drm);
@@ -1554,6 +1484,8 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 
 #define GLK_REVID_A0		0x0
 #define GLK_REVID_A1		0x1
+#define GLK_REVID_A2		0x2
+#define GLK_REVID_B0		0x3
 
 #define IS_GLK_REVID(dev_priv, since, until) \
 	(IS_GEMINILAKE(dev_priv) && IS_REVID(dev_priv, since, until))
@@ -1737,11 +1669,6 @@ intel_ggtt_update_needs_vtd_wa(struct drm_i915_private *dev_priv)
 }
 
 /* i915_drv.c */
-#ifdef CONFIG_COMPAT
-long i915_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg);
-#else
-#define i915_compat_ioctl NULL
-#endif
 extern const struct dev_pm_ops i915_pm_ops;
 
 int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
@@ -1750,16 +1677,6 @@ void i915_driver_remove(struct drm_i915_private *i915);
 int i915_resume_switcheroo(struct drm_i915_private *i915);
 int i915_suspend_switcheroo(struct drm_i915_private *i915, pm_message_t state);
 
-static inline bool intel_gvt_active(struct drm_i915_private *dev_priv)
-{
-	return dev_priv->gvt;
-}
-
-static inline bool intel_vgpu_active(struct drm_i915_private *dev_priv)
-{
-	return dev_priv->vgpu.active;
-}
-
 int i915_getparam_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 
@@ -1824,12 +1741,6 @@ int i915_gem_object_unbind(struct drm_i915_gem_object *obj,
 
 void i915_gem_runtime_suspend(struct drm_i915_private *dev_priv);
 
-static inline int __must_check
-i915_mutex_lock_interruptible(struct drm_device *dev)
-{
-	return mutex_lock_interruptible(&dev->struct_mutex);
-}
-
 int i915_gem_dumb_create(struct drm_file *file_priv,
 			 struct drm_device *dev,
 			 struct drm_mode_create_dumb *args);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index d92cf966fa3f..ca5420012a22 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -26,7 +26,6 @@
  */
 
 #include <drm/drm_vma_manager.h>
-#include <drm/i915_drm.h>
 #include <linux/dma-fence-array.h>
 #include <linux/kthread.h>
 #include <linux/dma-resv.h>
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 0697bedebeef..4518b9b35c3d 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -26,8 +26,6 @@
  *
  */
 
-#include <drm/i915_drm.h>
-
 #include "gem/i915_gem_context.h"
 #include "gt/intel_gt_requests.h"
 
@@ -292,7 +290,8 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
 		GEM_BUG_ON(!drm_mm_node_allocated(node));
 		vma = container_of(node, typeof(*vma), node);
 
-		/* If we are using coloring to insert guard pages between
+		/*
+		 * If we are using coloring to insert guard pages between
 		 * different cache domains within the address space, we have
 		 * to check whether the objects on either side of our range
 		 * abutt and conflict. If they are in conflict, then we evict
@@ -309,22 +308,18 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
 			}
 		}
 
-		if (flags & PIN_NONBLOCK &&
-		    (i915_vma_is_pinned(vma) || i915_vma_is_active(vma))) {
+		if (i915_vma_is_pinned(vma)) {
 			ret = -ENOSPC;
 			break;
 		}
 
-		/* Overlap of objects in the same batch? */
-		if (i915_vma_is_pinned(vma)) {
+		if (flags & PIN_NONBLOCK && i915_vma_is_active(vma)) {
 			ret = -ENOSPC;
-			if (vma->exec_flags &&
-			    *vma->exec_flags & EXEC_OBJECT_PINNED)
-				ret = -EINVAL;
 			break;
 		}
 
-		/* Never show fear in the face of dragons!
+		/*
+		 * Never show fear in the face of dragons!
 		 *
 		 * We cannot directly remove this node from within this
 		 * iterator and as with i915_gem_evict_something() we employ
diff --git a/drivers/gpu/drm/i915/i915_gem_fence_reg.c b/drivers/gpu/drm/i915/i915_gem_fence_reg.c
index 049cd3785347..d152b648c73c 100644
--- a/drivers/gpu/drm/i915/i915_gem_fence_reg.c
+++ b/drivers/gpu/drm/i915/i915_gem_fence_reg.c
@@ -21,10 +21,9 @@
  * IN THE SOFTWARE.
  */
 
-#include <drm/i915_drm.h>
-
 #include "i915_drv.h"
 #include "i915_scatterlist.h"
+#include "i915_pvinfo.h"
 #include "i915_vgpu.h"
 
 /**
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index e7834fa1e0ac..cb43381b0d37 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -15,8 +15,6 @@
 #include <asm/set_memory.h>
 #include <asm/smp.h>
 
-#include <drm/i915_drm.h>
-
 #include "display/intel_frontbuffer.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_requests.h"
diff --git a/drivers/gpu/drm/i915/i915_ioc32.c b/drivers/gpu/drm/i915/i915_ioc32.c
index c1007245f46d..8e45ca3d2ede 100644
--- a/drivers/gpu/drm/i915/i915_ioc32.c
+++ b/drivers/gpu/drm/i915/i915_ioc32.c
@@ -28,9 +28,10 @@
  */
 #include <linux/compat.h>
 
-#include <drm/i915_drm.h>
 #include <drm/drm_ioctl.h>
+
 #include "i915_drv.h"
+#include "i915_ioc32.h"
 
 struct drm_i915_getparam32 {
 	s32 param;
@@ -67,7 +68,7 @@ static drm_ioctl_compat_t *i915_compat_ioctls[] = {
 };
 
 /**
- * i915_compat_ioctl - handle the mistakes of the past
+ * i915_ioc32_compat_ioctl - handle the mistakes of the past
  * @filp: the file pointer
  * @cmd: the ioctl command (and encoded flags)
  * @arg: the ioctl argument (from userspace)
@@ -75,7 +76,7 @@ static drm_ioctl_compat_t *i915_compat_ioctls[] = {
  * Called whenever a 32-bit process running under a 64-bit kernel
  * performs an ioctl on /dev/dri/card<n>.
  */
-long i915_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
+long i915_ioc32_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 {
 	unsigned int nr = DRM_IOCTL_NR(cmd);
 	drm_ioctl_compat_t *fn = NULL;
diff --git a/drivers/gpu/drm/i915/i915_ioc32.h b/drivers/gpu/drm/i915/i915_ioc32.h
new file mode 100644
index 000000000000..40dcd55ca213
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_ioc32.h
@@ -0,0 +1,17 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2020 Intel Corporation
+ */
+
+#ifndef __I915_IOC32_H__
+#define __I915_IOC32_H__
+
+#ifdef CONFIG_COMPAT
+struct file;
+long i915_ioc32_compat_ioctl(struct file *filp, unsigned int cmd,
+			     unsigned long arg);
+#else
+#define i915_ioc32_compat_ioctl NULL
+#endif
+
+#endif /* __I915_IOC32_H__ */
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 5d1b11346145..9f0653cf0510 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -34,7 +34,6 @@
 
 #include <drm/drm_drv.h>
 #include <drm/drm_irq.h>
-#include <drm/i915_drm.h>
 
 #include "display/intel_display_types.h"
 #include "display/intel_fifo_underrun.h"
@@ -169,6 +168,14 @@ static const u32 hpd_tgp[HPD_NUM_PINS] = {
 	[HPD_PORT_I] = SDE_TC_HOTPLUG_ICP(PORT_TC6),
 };
 
+static void
+intel_handle_vblank(struct drm_i915_private *dev_priv, enum pipe pipe)
+{
+	struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
+
+	drm_crtc_handle_vblank(&crtc->base);
+}
+
 void gen3_irq_reset(struct intel_uncore *uncore, i915_reg_t imr,
 		    i915_reg_t iir, i915_reg_t ier)
 {
@@ -1210,8 +1217,8 @@ static void display_pipe_crc_irq_handler(struct drm_i915_private *dev_priv,
 					 u32 crc2, u32 crc3,
 					 u32 crc4)
 {
-	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
 	struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
+	struct intel_pipe_crc *pipe_crc = &crtc->pipe_crc;
 	u32 crcs[5] = { crc0, crc1, crc2, crc3, crc4 };
 
 	trace_intel_pipe_crc(crtc, crcs);
@@ -1374,7 +1381,7 @@ static void i8xx_pipestat_irq_handler(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS)
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 
 		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 			i9xx_pipe_crc_irq_handler(dev_priv, pipe);
@@ -1392,7 +1399,7 @@ static void i915_pipestat_irq_handler(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS)
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 
 		if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS)
 			blc_event = true;
@@ -1416,7 +1423,7 @@ static void i965_pipestat_irq_handler(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 
 		if (pipe_stats[pipe] & PIPE_LEGACY_BLC_EVENT_STATUS)
 			blc_event = true;
@@ -1442,7 +1449,7 @@ static void valleyview_pipestat_irq_handler(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 
 		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 			i9xx_pipe_crc_irq_handler(dev_priv, pipe);
@@ -1749,11 +1756,12 @@ static void ibx_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
 	if (pch_iir & SDE_POISON)
 		drm_err(&dev_priv->drm, "PCH poison interrupt\n");
 
-	if (pch_iir & SDE_FDI_MASK)
+	if (pch_iir & SDE_FDI_MASK) {
 		for_each_pipe(dev_priv, pipe)
 			drm_dbg(&dev_priv->drm, "  pipe %c FDI IIR: 0x%08x\n",
 				pipe_name(pipe),
 				I915_READ(FDI_RX_IIR(pipe)));
+	}
 
 	if (pch_iir & (SDE_TRANSB_CRC_DONE | SDE_TRANSA_CRC_DONE))
 		drm_dbg(&dev_priv->drm, "PCH transcoder CRC done interrupt\n");
@@ -1833,11 +1841,12 @@ static void cpt_irq_handler(struct drm_i915_private *dev_priv, u32 pch_iir)
 	if (pch_iir & SDE_AUDIO_CP_CHG_CPT)
 		drm_dbg(&dev_priv->drm, "Audio CP change interrupt\n");
 
-	if (pch_iir & SDE_FDI_MASK_CPT)
+	if (pch_iir & SDE_FDI_MASK_CPT) {
 		for_each_pipe(dev_priv, pipe)
 			drm_dbg(&dev_priv->drm, "  pipe %c FDI IIR: 0x%08x\n",
 				pipe_name(pipe),
 				I915_READ(FDI_RX_IIR(pipe)));
+	}
 
 	if (pch_iir & SDE_ERROR_CPT)
 		cpt_serr_int_handler(dev_priv);
@@ -1978,7 +1987,7 @@ static void ilk_display_irq_handler(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		if (de_iir & DE_PIPE_VBLANK(pipe))
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 
 		if (de_iir & DE_PIPE_FIFO_UNDERRUN(pipe))
 			intel_cpu_fifo_underrun_irq_handler(dev_priv, pipe);
@@ -2031,7 +2040,7 @@ static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
 
 	for_each_pipe(dev_priv, pipe) {
 		if (de_iir & (DE_PIPE_VBLANK_IVB(pipe)))
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 	}
 
 	/* check event from PCH */
@@ -2344,7 +2353,7 @@ gen8_de_irq_handler(struct drm_i915_private *dev_priv, u32 master_ctl)
 		I915_WRITE(GEN8_DE_PIPE_IIR(pipe), iir);
 
 		if (iir & GEN8_PIPE_VBLANK)
-			drm_handle_vblank(&dev_priv->drm, pipe);
+			intel_handle_vblank(dev_priv, pipe);
 
 		if (iir & GEN8_PIPE_CDCLK_CRC_DONE)
 			hsw_pipe_crc_irq_handler(dev_priv, pipe);
diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
index 1d678aa7d420..2c80a0194c80 100644
--- a/drivers/gpu/drm/i915/i915_pci.c
+++ b/drivers/gpu/drm/i915/i915_pci.c
@@ -26,6 +26,7 @@
 #include <linux/vga_switcheroo.h>
 
 #include <drm/drm_drv.h>
+#include <drm/i915_pciids.h>
 
 #include "display/intel_fbdev.h"
 
@@ -822,7 +823,6 @@ static const struct intel_device_info tgl_info = {
 	GEN12_FEATURES,
 	PLATFORM(INTEL_TIGERLAKE),
 	.pipe_mask = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C) | BIT(PIPE_D),
-	.require_force_probe = 1,
 	.display.has_modular_fia = 1,
 	.engine_mask =
 		BIT(RCS0) | BIT(BCS0) | BIT(VECS0) | BIT(VCS0) | BIT(VCS2),
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 79391d92ab7e..1b074bb4a7fe 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -1405,8 +1405,10 @@ static void i915_oa_stream_destroy(struct i915_perf_stream *stream)
 	/*
 	 * Unset exclusive_stream first, it will be checked while disabling
 	 * the metric set on gen8+.
+	 *
+	 * See i915_oa_init_reg_state() and lrc_configure_all_contexts()
 	 */
-	perf->exclusive_stream = NULL;
+	WRITE_ONCE(perf->exclusive_stream, NULL);
 	perf->ops.disable_metric_set(stream);
 
 	free_oa_buffer(stream);
@@ -2200,7 +2202,9 @@ static int gen8_modify_self(struct intel_context *ce,
 	struct i915_request *rq;
 	int err;
 
+	intel_engine_pm_get(ce->engine);
 	rq = i915_request_create(ce);
+	intel_engine_pm_put(ce->engine);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
@@ -2869,7 +2873,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 		goto err_oa_buf_alloc;
 
 	stream->ops = &i915_oa_stream_ops;
-	perf->exclusive_stream = stream;
+	WRITE_ONCE(perf->exclusive_stream, stream);
 
 	ret = i915_perf_stream_enable_sync(stream);
 	if (ret) {
@@ -2889,7 +2893,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	return 0;
 
 err_enable:
-	perf->exclusive_stream = NULL;
+	WRITE_ONCE(perf->exclusive_stream, NULL);
 	perf->ops.disable_metric_set(stream);
 
 	free_oa_buffer(stream);
@@ -2915,12 +2919,11 @@ void i915_oa_init_reg_state(const struct intel_context *ce,
 {
 	struct i915_perf_stream *stream;
 
-	/* perf.exclusive_stream serialised by lrc_configure_all_contexts() */
-
 	if (engine->class != RENDER_CLASS)
 		return;
 
-	stream = engine->i915->perf.exclusive_stream;
+	/* perf.exclusive_stream serialised by lrc_configure_all_contexts() */
+	stream = READ_ONCE(engine->i915->perf.exclusive_stream);
 	/*
 	 * For gen12, only CTX_R_PWR_CLK_STATE needs update, but the caller
 	 * is already doing that, so nothing to be done for gen12 here.
diff --git a/drivers/gpu/drm/i915/i915_pmu.h b/drivers/gpu/drm/i915/i915_pmu.h
index f1d6cad0d7d5..941f0c14037c 100644
--- a/drivers/gpu/drm/i915/i915_pmu.h
+++ b/drivers/gpu/drm/i915/i915_pmu.h
@@ -10,7 +10,7 @@
 #include <linux/hrtimer.h>
 #include <linux/perf_event.h>
 #include <linux/spinlock_types.h>
-#include <drm/i915_drm.h>
+#include <uapi/drm/i915_drm.h>
 
 struct drm_i915_private;
 
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index f45b5e86ec63..309cb7d96b35 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -3285,6 +3285,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 
 /* Framebuffer compression for Ivybridge */
 #define IVB_FBC_RT_BASE			_MMIO(0x7020)
+#define IVB_FBC_RT_BASE_UPPER		_MMIO(0x7024)
 
 #define IPS_CTL		_MMIO(0x43408)
 #define   IPS_ENABLE	(1 << 31)
@@ -4860,16 +4861,6 @@ enum {
 #define _PP_STATUS			0x61200
 #define PP_STATUS(pps_idx)		_MMIO_PPS(pps_idx, _PP_STATUS)
 #define   PP_ON				REG_BIT(31)
-
-#define _PP_CONTROL_1			0xc7204
-#define _PP_CONTROL_2			0xc7304
-#define ICP_PP_CONTROL(x)		_MMIO(((x) == 1) ? _PP_CONTROL_1 : \
-					      _PP_CONTROL_2)
-#define  POWER_CYCLE_DELAY_MASK		REG_GENMASK(8, 4)
-#define  VDD_OVERRIDE_FORCE		REG_BIT(3)
-#define  BACKLIGHT_ENABLE		REG_BIT(2)
-#define  PWR_DOWN_ON_RESET		REG_BIT(1)
-#define  PWR_STATE_TARGET		REG_BIT(0)
 /*
  * Indicates that all dependencies of the panel are on:
  *
@@ -5880,7 +5871,6 @@ enum {
 
 #define  _PIPEAGCMAX           0x70010
 #define  _PIPEBGCMAX           0x71010
-#define PIPEGCMAX_RGB_MASK     REG_GENMASK(15, 0)
 #define PIPEGCMAX(pipe, i)     _MMIO_PIPE2(pipe, _PIPEAGCMAX + (i) * 4)
 
 #define _PIPE_MISC_A			0x70030
@@ -5889,6 +5879,7 @@ enum {
 #define   PIPEMISC_YUV420_MODE_FULL_BLEND (1 << 26) /* glk+ */
 #define   PIPEMISC_HDR_MODE_PRECISION	(1 << 23) /* icl+ */
 #define   PIPEMISC_OUTPUT_COLORSPACE_YUV  (1 << 11)
+#define   PIPEMISC_PIXEL_ROUNDING_TRUNC	REG_BIT(8) /* tgl+ */
 #define   PIPEMISC_DITHER_BPC_MASK	(7 << 5)
 #define   PIPEMISC_DITHER_8_BPC		(0 << 5)
 #define   PIPEMISC_DITHER_10_BPC	(1 << 5)
@@ -9149,14 +9140,19 @@ enum {
 #define   THROTTLE_12_5				(7 << 2)
 #define   DISABLE_EARLY_EOT			(1 << 1)
 
-#define GEN7_ROW_CHICKEN2		_MMIO(0xe4f4)
-#define GEN12_DISABLE_EARLY_READ	BIT(14)
+#define GEN7_ROW_CHICKEN2			_MMIO(0xe4f4)
+#define   GEN12_DISABLE_EARLY_READ		REG_BIT(14)
+#define   GEN12_PUSH_CONST_DEREF_HOLD_DIS	REG_BIT(8)
 
 #define GEN7_ROW_CHICKEN2_GT2		_MMIO(0xf4f4)
 #define   DOP_CLOCK_GATING_DISABLE	(1 << 0)
 #define   PUSH_CONSTANT_DEREF_DISABLE	(1 << 8)
 #define   GEN11_TDL_CLOCK_GATING_FIX_DISABLE	(1 << 1)
 
+#define GEN9_ROW_CHICKEN4		_MMIO(0xe48c)
+#define   GEN12_DISABLE_TDL_PUSH	REG_BIT(9)
+#define   GEN11_DIS_PICK_2ND_EU		REG_BIT(7)
+
 #define HSW_ROW_CHICKEN3		_MMIO(0xe49c)
 #define  HSW_ROW_CHICKEN3_L3_GLOBAL_ATOMICS_DISABLE    (1 << 6)
 
@@ -9257,6 +9253,10 @@ enum {
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI_74250	(7 << 16)
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI_148352	(8 << 16)
 #define   AUD_CONFIG_PIXEL_CLOCK_HDMI_148500	(9 << 16)
+#define   AUD_CONFIG_PIXEL_CLOCK_HDMI_296703	(10 << 16)
+#define   AUD_CONFIG_PIXEL_CLOCK_HDMI_297000	(11 << 16)
+#define   AUD_CONFIG_PIXEL_CLOCK_HDMI_593407	(12 << 16)
+#define   AUD_CONFIG_PIXEL_CLOCK_HDMI_594000	(13 << 16)
 #define   AUD_CONFIG_DISABLE_NCTS		(1 << 3)
 
 /* HSW Audio */
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index e5a55801f753..c0df71d7d0ff 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -290,7 +290,7 @@ bool i915_request_retire(struct i915_request *rq)
 	spin_unlock_irq(&rq->lock);
 
 	remove_from_client(rq);
-	list_del_rcu(&rq->link);
+	__list_del_entry(&rq->link); /* poison neither prev/next (RCU walks) */
 
 	intel_context_exit(rq->context);
 	intel_context_unpin(rq->context);
@@ -363,6 +363,50 @@ __await_execution(struct i915_request *rq,
 	return 0;
 }
 
+static bool fatal_error(int error)
+{
+	switch (error) {
+	case 0: /* not an error! */
+	case -EAGAIN: /* innocent victim of a GT reset (__i915_request_reset) */
+	case -ETIMEDOUT: /* waiting for Godot (timer_i915_sw_fence_wake) */
+		return false;
+	default:
+		return true;
+	}
+}
+
+void __i915_request_skip(struct i915_request *rq)
+{
+	GEM_BUG_ON(!fatal_error(rq->fence.error));
+
+	if (rq->infix == rq->postfix)
+		return;
+
+	/*
+	 * As this request likely depends on state from the lost
+	 * context, clear out all the user operations leaving the
+	 * breadcrumb at the end (so we get the fence notifications).
+	 */
+	__i915_request_fill(rq, 0);
+	rq->infix = rq->postfix;
+}
+
+void i915_request_set_error_once(struct i915_request *rq, int error)
+{
+	int old;
+
+	GEM_BUG_ON(!IS_ERR_VALUE((long)error));
+
+	if (i915_request_signaled(rq))
+		return;
+
+	old = READ_ONCE(rq->fence.error);
+	do {
+		if (fatal_error(old))
+			return;
+	} while (!try_cmpxchg(&rq->fence.error, &old, error));
+}
+
 bool __i915_request_submit(struct i915_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
@@ -392,8 +436,10 @@ bool __i915_request_submit(struct i915_request *request)
 	if (i915_request_completed(request))
 		goto xfer;
 
-	if (intel_context_is_banned(request->context))
-		i915_request_skip(request, -EIO);
+	if (unlikely(intel_context_is_banned(request->context)))
+		i915_request_set_error_once(request, -EIO);
+	if (unlikely(fatal_error(request->fence.error)))
+		__i915_request_skip(request);
 
 	/*
 	 * Are we using semaphores when the gpu is already saturated?
@@ -519,7 +565,7 @@ submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
 		trace_i915_request_submit(request);
 
 		if (unlikely(fence->error))
-			i915_request_skip(request, fence->error);
+			i915_request_set_error_once(request, fence->error);
 
 		/*
 		 * We need to serialize use of the submit_request() callback
@@ -542,19 +588,31 @@ submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
 	return NOTIFY_DONE;
 }
 
+static void irq_semaphore_cb(struct irq_work *wrk)
+{
+	struct i915_request *rq =
+		container_of(wrk, typeof(*rq), semaphore_work);
+
+	i915_schedule_bump_priority(rq, I915_PRIORITY_NOSEMAPHORE);
+	i915_request_put(rq);
+}
+
 static int __i915_sw_fence_call
 semaphore_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
 {
-	struct i915_request *request =
-		container_of(fence, typeof(*request), semaphore);
+	struct i915_request *rq = container_of(fence, typeof(*rq), semaphore);
 
 	switch (state) {
 	case FENCE_COMPLETE:
-		i915_schedule_bump_priority(request, I915_PRIORITY_NOSEMAPHORE);
+		if (!(READ_ONCE(rq->sched.attr.priority) & I915_PRIORITY_NOSEMAPHORE)) {
+			i915_request_get(rq);
+			init_irq_work(&rq->semaphore_work, irq_semaphore_cb);
+			irq_work_queue(&rq->semaphore_work);
+		}
 		break;
 
 	case FENCE_FREE:
-		i915_request_put(request);
+		i915_request_put(rq);
 		break;
 	}
 
@@ -691,6 +749,7 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 	RCU_INIT_POINTER(rq->timeline, tl);
 	RCU_INIT_POINTER(rq->hwsp_cacheline, tl->hwsp_cacheline);
 	rq->hwsp_seqno = tl->hwsp_seqno;
+	GEM_BUG_ON(i915_request_completed(rq));
 
 	rq->rcustate = get_state_synchronize_rcu(); /* acts as smp_mb() */
 
@@ -791,8 +850,8 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 	struct dma_fence *fence;
 	int err;
 
-	GEM_BUG_ON(i915_request_timeline(rq) ==
-		   rcu_access_pointer(signal->timeline));
+	if (i915_request_timeline(rq) == rcu_access_pointer(signal->timeline))
+		return 0;
 
 	if (i915_request_started(signal))
 		return 0;
@@ -836,7 +895,7 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 		return 0;
 
 	err = 0;
-	if (intel_timeline_sync_is_later(i915_request_timeline(rq), fence))
+	if (!intel_timeline_sync_is_later(i915_request_timeline(rq), fence))
 		err = i915_sw_fence_await_dma_fence(&rq->submit,
 						    fence, 0,
 						    I915_FENCE_GFP);
@@ -860,7 +919,7 @@ already_busywaiting(struct i915_request *rq)
 	 *
 	 * See the are-we-too-late? check in __i915_request_submit().
 	 */
-	return rq->sched.semaphores | rq->engine->saturated;
+	return rq->sched.semaphores | READ_ONCE(rq->engine->saturated);
 }
 
 static int
@@ -918,6 +977,8 @@ emit_semaphore_wait(struct i915_request *to,
 		    struct i915_request *from,
 		    gfp_t gfp)
 {
+	const intel_engine_mask_t mask = READ_ONCE(from->engine)->mask;
+
 	if (!intel_context_use_semaphores(to->context))
 		goto await_fence;
 
@@ -925,7 +986,7 @@ emit_semaphore_wait(struct i915_request *to,
 		goto await_fence;
 
 	/* Just emit the first semaphore we see as request space is limited. */
-	if (already_busywaiting(to) & from->engine->mask)
+	if (already_busywaiting(to) & mask)
 		goto await_fence;
 
 	if (i915_request_await_start(to, from) < 0)
@@ -938,7 +999,7 @@ emit_semaphore_wait(struct i915_request *to,
 	if (__emit_semaphore_wait(to, from, from->fence.seqno))
 		goto await_fence;
 
-	to->sched.semaphores |= from->engine->mask;
+	to->sched.semaphores |= mask;
 	to->sched.flags |= I915_SCHED_HAS_SEMAPHORE_CHAIN;
 	return 0;
 
@@ -1081,14 +1142,45 @@ __i915_request_await_execution(struct i915_request *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);
+	/*
+	 * Wait until the start of this request.
+	 *
+	 * The execution cb fires when we submit the request to HW. But in
+	 * many cases this may be long before the request itself is ready to
+	 * run (consider that we submit 2 requests for the same context, where
+	 * the request of interest is behind an indefinite spinner). So we hook
+	 * up to both to reduce our queues and keep the execution lag minimised
+	 * in the worst case, though we hope that the await_start is elided.
+	 */
+	err = i915_request_await_start(to, from);
 	if (err < 0)
 		return err;
 
+	/*
+	 * Ensure both start together [after all semaphores in signal]
+	 *
+	 * Now that we are queued to the HW at roughly the same time (thanks
+	 * to the execute cb) and are ready to run at roughly the same time
+	 * (thanks to the await start), our signaler may still be indefinitely
+	 * delayed by waiting on a semaphore from a remote engine. If our
+	 * signaler depends on a semaphore, so indirectly do we, and we do not
+	 * want to start our payload until our signaler also starts theirs.
+	 * So we wait.
+	 *
+	 * However, there is also a second condition for which we need to wait
+	 * for the precise start of the signaler. Consider that the signaler
+	 * was submitted in a chain of requests following another context
+	 * (with just an ordinary intra-engine fence dependency between the
+	 * two). In this case the signaler is queued to HW, but not for
+	 * immediate execution, and so we must wait until it reaches the
+	 * active slot.
+	 */
+	if (intel_engine_has_semaphores(to->engine)) {
+		err = __emit_semaphore_wait(to, from, from->fence.seqno - 1);
+		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);
@@ -1209,23 +1301,6 @@ i915_request_await_object(struct i915_request *to,
 	return ret;
 }
 
-void i915_request_skip(struct i915_request *rq, int error)
-{
-	GEM_BUG_ON(!IS_ERR_VALUE((long)error));
-	dma_fence_set_error(&rq->fence, error);
-
-	if (rq->infix == rq->postfix)
-		return;
-
-	/*
-	 * As this request likely depends on state from the lost
-	 * context, clear out all the user operations leaving the
-	 * breadcrumb at the end (so we get the fence notifications).
-	 */
-	__i915_request_fill(rq, 0);
-	rq->infix = rq->postfix;
-}
-
 static struct i915_request *
 __i915_request_add_to_timeline(struct i915_request *rq)
 {
@@ -1255,7 +1330,17 @@ __i915_request_add_to_timeline(struct i915_request *rq)
 	prev = to_request(__i915_active_fence_set(&timeline->last_request,
 						  &rq->fence));
 	if (prev && !i915_request_completed(prev)) {
-		if (is_power_of_2(prev->engine->mask | rq->engine->mask))
+		/*
+		 * The requests are supposed to be kept in order. However,
+		 * we need to be wary in case the timeline->last_request
+		 * is used as a barrier for external modification to this
+		 * context.
+		 */
+		GEM_BUG_ON(prev->context == rq->context &&
+			   i915_seqno_passed(prev->fence.seqno,
+					     rq->fence.seqno));
+
+		if (is_power_of_2(READ_ONCE(prev->engine)->mask | rq->engine->mask))
 			i915_sw_fence_await_sw_fence(&rq->submit,
 						     &prev->submit,
 						     &rq->submitq);
@@ -1329,9 +1414,9 @@ void __i915_request_queue(struct i915_request *rq,
 	 * decide whether to preempt the entire chain so that it is ready to
 	 * run at the earliest possible convenience.
 	 */
-	i915_sw_fence_commit(&rq->semaphore);
 	if (attr && rq->engine->schedule)
 		rq->engine->schedule(rq, attr);
+	i915_sw_fence_commit(&rq->semaphore);
 	i915_sw_fence_commit(&rq->submit);
 }
 
@@ -1339,39 +1424,23 @@ void i915_request_add(struct i915_request *rq)
 {
 	struct intel_timeline * const tl = i915_request_timeline(rq);
 	struct i915_sched_attr attr = {};
-	struct i915_request *prev;
+	struct i915_gem_context *ctx;
 
 	lockdep_assert_held(&tl->mutex);
 	lockdep_unpin_lock(&tl->mutex, rq->cookie);
 
 	trace_i915_request_add(rq);
+	__i915_request_commit(rq);
 
-	prev = __i915_request_commit(rq);
-
-	if (rcu_access_pointer(rq->context->gem_context))
-		attr = i915_request_gem_context(rq)->sched;
+	/* XXX placeholder for selftests */
+	rcu_read_lock();
+	ctx = rcu_dereference(rq->context->gem_context);
+	if (ctx)
+		attr = ctx->sched;
+	rcu_read_unlock();
 
-	/*
-	 * Boost actual workloads past semaphores!
-	 *
-	 * With semaphores we spin on one engine waiting for another,
-	 * simply to reduce the latency of starting our work when
-	 * the signaler completes. However, if there is any other
-	 * work that we could be doing on this engine instead, that
-	 * is better utilisation and will reduce the overall duration
-	 * of the current work. To avoid PI boosting a semaphore
-	 * far in the distance past over useful work, we keep a history
-	 * of any semaphore use along our dependency chain.
-	 */
 	if (!(rq->sched.flags & I915_SCHED_HAS_SEMAPHORE_CHAIN))
 		attr.priority |= I915_PRIORITY_NOSEMAPHORE;
-
-	/*
-	 * Boost priorities to new clients (new request flows).
-	 *
-	 * Allow interactive/synchronous clients to jump ahead of
-	 * the bulk clients. (FQ_CODEL)
-	 */
 	if (list_empty(&rq->sched.signalers_list))
 		attr.priority |= I915_PRIORITY_WAIT;
 
@@ -1379,32 +1448,10 @@ void i915_request_add(struct i915_request *rq)
 	__i915_request_queue(rq, &attr);
 	local_bh_enable(); /* Kick the execlists tasklet if just scheduled */
 
-	/*
-	 * In typical scenarios, we do not expect the previous request on
-	 * the timeline to be still tracked by timeline->last_request if it
-	 * has been completed. If the completed request is still here, that
-	 * implies that request retirement is a long way behind submission,
-	 * suggesting that we haven't been retiring frequently enough from
-	 * the combination of retire-before-alloc, waiters and the background
-	 * retirement worker. So if the last request on this timeline was
-	 * already completed, do a catch up pass, flushing the retirement queue
-	 * up to this client. Since we have now moved the heaviest operations
-	 * during retirement onto secondary workers, such as freeing objects
-	 * or contexts, retiring a bunch of requests is mostly list management
-	 * (and cache misses), and so we should not be overly penalizing this
-	 * client by performing excess work, though we may still performing
-	 * work on behalf of others -- but instead we should benefit from
-	 * improved resource management. (Well, that's the theory at least.)
-	 */
-	if (prev &&
-	    i915_request_completed(prev) &&
-	    rcu_access_pointer(prev->timeline) == tl)
-		i915_request_retire_upto(prev);
-
 	mutex_unlock(&tl->mutex);
 }
 
-static unsigned long local_clock_us(unsigned int *cpu)
+static unsigned long local_clock_ns(unsigned int *cpu)
 {
 	unsigned long t;
 
@@ -1421,7 +1468,7 @@ static unsigned long local_clock_us(unsigned int *cpu)
 	 * stop busywaiting, see busywait_stop().
 	 */
 	*cpu = get_cpu();
-	t = local_clock() >> 10;
+	t = local_clock();
 	put_cpu();
 
 	return t;
@@ -1431,15 +1478,15 @@ static bool busywait_stop(unsigned long timeout, unsigned int cpu)
 {
 	unsigned int this_cpu;
 
-	if (time_after(local_clock_us(&this_cpu), timeout))
+	if (time_after(local_clock_ns(&this_cpu), timeout))
 		return true;
 
 	return this_cpu != cpu;
 }
 
-static bool __i915_spin_request(const struct i915_request * const rq,
-				int state, unsigned long timeout_us)
+static bool __i915_spin_request(const struct i915_request * const rq, int state)
 {
+	unsigned long timeout_ns;
 	unsigned int cpu;
 
 	/*
@@ -1467,7 +1514,8 @@ static bool __i915_spin_request(const struct i915_request * const rq,
 	 * takes to sleep on a request, on the order of a microsecond.
 	 */
 
-	timeout_us += local_clock_us(&cpu);
+	timeout_ns = READ_ONCE(rq->engine->props.max_busywait_duration_ns);
+	timeout_ns += local_clock_ns(&cpu);
 	do {
 		if (i915_request_completed(rq))
 			return true;
@@ -1475,7 +1523,7 @@ static bool __i915_spin_request(const struct i915_request * const rq,
 		if (signal_pending_state(state, current))
 			break;
 
-		if (busywait_stop(timeout_us, cpu))
+		if (busywait_stop(timeout_ns, cpu))
 			break;
 
 		cpu_relax();
@@ -1561,8 +1609,8 @@ long i915_request_wait(struct i915_request *rq,
 	 * completion. That requires having a good predictor for the request
 	 * duration, which we currently lack.
 	 */
-	if (IS_ACTIVE(CONFIG_DRM_I915_SPIN_REQUEST) &&
-	    __i915_spin_request(rq, state, CONFIG_DRM_I915_SPIN_REQUEST)) {
+	if (IS_ACTIVE(CONFIG_DRM_I915_MAX_REQUEST_BUSYWAIT) &&
+	    __i915_spin_request(rq, state)) {
 		dma_fence_signal(&rq->fence);
 		goto out;
 	}
diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index da8420f03232..3c552bfea67a 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -26,6 +26,7 @@
 #define I915_REQUEST_H
 
 #include <linux/dma-fence.h>
+#include <linux/irq_work.h>
 #include <linux/lockdep.h>
 
 #include "gem/i915_gem_context_types.h"
@@ -208,6 +209,7 @@ struct i915_request {
 	};
 	struct list_head execute_cb;
 	struct i915_sw_fence semaphore;
+	struct irq_work semaphore_work;
 
 	/*
 	 * A list of everyone we wait upon, and everyone who waits upon us.
@@ -303,6 +305,9 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp);
 struct i915_request * __must_check
 i915_request_create(struct intel_context *ce);
 
+void i915_request_set_error_once(struct i915_request *rq, int error);
+void __i915_request_skip(struct i915_request *rq);
+
 struct i915_request *__i915_request_commit(struct i915_request *request);
 void __i915_request_queue(struct i915_request *rq,
 			  const struct i915_sched_attr *attr);
@@ -352,8 +357,6 @@ void i915_request_add(struct i915_request *rq);
 bool __i915_request_submit(struct i915_request *request);
 void i915_request_submit(struct i915_request *request);
 
-void i915_request_skip(struct i915_request *request, int error);
-
 void __i915_request_unsubmit(struct i915_request *request);
 void i915_request_unsubmit(struct i915_request *request);
 
@@ -395,7 +398,9 @@ static inline bool i915_seqno_passed(u32 seq1, u32 seq2)
 
 static inline u32 __hwsp_seqno(const struct i915_request *rq)
 {
-	return READ_ONCE(*rq->hwsp_seqno);
+	const u32 *hwsp = READ_ONCE(rq->hwsp_seqno);
+
+	return READ_ONCE(*hwsp);
 }
 
 /**
@@ -509,7 +514,8 @@ static inline bool i915_request_completed(const struct i915_request *rq)
 
 static inline void i915_request_mark_complete(struct i915_request *rq)
 {
-	rq->hwsp_seqno = (u32 *)&rq->fence.seqno; /* decouple from HWSP */
+	WRITE_ONCE(rq->hwsp_seqno, /* decouple from HWSP */
+		   (u32 *)&rq->fence.seqno);
 }
 
 static inline bool i915_request_has_waitboost(const struct i915_request *rq)
diff --git a/drivers/gpu/drm/i915/i915_scheduler.c b/drivers/gpu/drm/i915/i915_scheduler.c
index be770f2419b1..68b06a7ba667 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.c
+++ b/drivers/gpu/drm/i915/i915_scheduler.c
@@ -209,6 +209,8 @@ static void kick_submission(struct intel_engine_cs *engine,
 	if (!inflight)
 		goto unlock;
 
+	engine->execlists.queue_priority_hint = prio;
+
 	/*
 	 * If we are already the currently executing context, don't
 	 * bother evaluating if we should preempt ourselves.
@@ -216,7 +218,6 @@ static void kick_submission(struct intel_engine_cs *engine,
 	if (inflight->context == rq->context)
 		goto unlock;
 
-	engine->execlists.queue_priority_hint = prio;
 	if (need_preempt(prio, rq_prio(inflight)))
 		tasklet_hi_schedule(&engine->execlists.tasklet);
 
@@ -227,10 +228,10 @@ unlock:
 static void __i915_schedule(struct i915_sched_node *node,
 			    const struct i915_sched_attr *attr)
 {
+	const int prio = max(attr->priority, node->attr.priority);
 	struct intel_engine_cs *engine;
 	struct i915_dependency *dep, *p;
 	struct i915_dependency stack;
-	const int prio = attr->priority;
 	struct sched_cache cache;
 	LIST_HEAD(dfs);
 
@@ -238,9 +239,6 @@ static void __i915_schedule(struct i915_sched_node *node,
 	lockdep_assert_held(&schedule_lock);
 	GEM_BUG_ON(prio == I915_PRIORITY_INVALID);
 
-	if (prio <= READ_ONCE(node->attr.priority))
-		return;
-
 	if (node_signaled(node))
 		return;
 
@@ -324,7 +322,7 @@ static void __i915_schedule(struct i915_sched_node *node,
 
 		GEM_BUG_ON(node_to_request(node)->engine != engine);
 
-		node->attr.priority = prio;
+		WRITE_ONCE(node->attr.priority, prio);
 
 		/*
 		 * Once the request is ready, it will be placed into the
diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c
index 8812cdd9007f..ed2be3489f8e 100644
--- a/drivers/gpu/drm/i915/i915_suspend.c
+++ b/drivers/gpu/drm/i915/i915_suspend.c
@@ -24,8 +24,6 @@
  * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <drm/i915_drm.h>
-
 #include "display/intel_fbc.h"
 #include "display/intel_gmbus.h"
 #include "display/intel_vga.h"
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index c14d762bd652..45d32ef42787 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -32,6 +32,7 @@
 
 #include "gt/intel_rc6.h"
 #include "gt/intel_rps.h"
+#include "gt/sysfs_engines.h"
 
 #include "i915_drv.h"
 #include "i915_sysfs.h"
@@ -606,6 +607,8 @@ void i915_setup_sysfs(struct drm_i915_private *dev_priv)
 		drm_err(&dev_priv->drm, "RPS sysfs setup failed\n");
 
 	i915_setup_error_capture(kdev);
+
+	intel_engines_add_sysfs(dev_priv);
 }
 
 void i915_teardown_sysfs(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/i915_utils.c b/drivers/gpu/drm/i915/i915_utils.c
index 632d6953c78d..029854ae65fc 100644
--- a/drivers/gpu/drm/i915/i915_utils.c
+++ b/drivers/gpu/drm/i915/i915_utils.c
@@ -8,7 +8,6 @@
 #include "i915_drv.h"
 #include "i915_utils.h"
 
-#define FDO_BUG_URL "https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs"
 #define FDO_BUG_MSG "Please file a bug on drm/i915; see " FDO_BUG_URL " for details."
 
 void
diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h
index b0ade76bec90..03a73d2bd50d 100644
--- a/drivers/gpu/drm/i915/i915_utils.h
+++ b/drivers/gpu/drm/i915/i915_utils.h
@@ -34,6 +34,8 @@
 struct drm_i915_private;
 struct timer_list;
 
+#define FDO_BUG_URL "https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs"
+
 #undef WARN_ON
 /* Many gcc seem to no see through this and fall over :( */
 #if 0
@@ -100,12 +102,24 @@ bool i915_error_injected(void);
 	typeof(max) max__ = (max); \
 	(void)(&start__ == &size__); \
 	(void)(&start__ == &max__); \
-	start__ > max__ || size__ > max__ - start__; \
+	start__ >= max__ || size__ > max__ - start__; \
 })
 
 #define range_overflows_t(type, start, size, max) \
 	range_overflows((type)(start), (type)(size), (type)(max))
 
+#define range_overflows_end(start, size, max) ({ \
+	typeof(start) start__ = (start); \
+	typeof(size) size__ = (size); \
+	typeof(max) max__ = (max); \
+	(void)(&start__ == &size__); \
+	(void)(&start__ == &max__); \
+	start__ > max__ || size__ > max__ - start__; \
+})
+
+#define range_overflows_end_t(type, start, size, max) \
+	range_overflows_end((type)(start), (type)(size), (type)(max))
+
 /* Note we don't consider signbits :| */
 #define overflows_type(x, T) \
 	(sizeof(x) > sizeof(T) && (x) >> BITS_PER_TYPE(T))
@@ -234,6 +248,11 @@ static inline u64 ptr_to_u64(const void *ptr)
 	__idx;								\
 })
 
+static inline bool is_power_of_2_u64(u64 n)
+{
+	return (n != 0 && ((n & (n - 1)) == 0));
+}
+
 static inline void __list_del_many(struct list_head *head,
 				   struct list_head *first)
 {
@@ -241,6 +260,12 @@ static inline void __list_del_many(struct list_head *head,
 	WRITE_ONCE(head->next, first);
 }
 
+static inline int list_is_last_rcu(const struct list_head *list,
+				   const struct list_head *head)
+{
+	return READ_ONCE(list->next) == head;
+}
+
 /*
  * Wait until the work is finally complete, even if it tries to postpone
  * by requeueing itself. Note, that if the worker never cancels itself,
diff --git a/drivers/gpu/drm/i915/i915_vgpu.c b/drivers/gpu/drm/i915/i915_vgpu.c
index 4afe21662266..70fca72f5162 100644
--- a/drivers/gpu/drm/i915/i915_vgpu.c
+++ b/drivers/gpu/drm/i915/i915_vgpu.c
@@ -21,6 +21,8 @@
  * SOFTWARE.
  */
 
+#include "i915_drv.h"
+#include "i915_pvinfo.h"
 #include "i915_vgpu.h"
 
 /**
@@ -51,13 +53,13 @@
  */
 
 /**
- * i915_detect_vgpu - detect virtual GPU
+ * intel_vgpu_detect - detect virtual GPU
  * @dev_priv: i915 device private
  *
  * This function is called at the initialization stage, to detect whether
  * running on a vGPU.
  */
-void i915_detect_vgpu(struct drm_i915_private *dev_priv)
+void intel_vgpu_detect(struct drm_i915_private *dev_priv)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	u64 magic;
@@ -102,11 +104,36 @@ out:
 	pci_iounmap(pdev, shared_area);
 }
 
+void intel_vgpu_register(struct drm_i915_private *i915)
+{
+	/*
+	 * Notify a valid surface after modesetting, when running inside a VM.
+	 */
+	if (intel_vgpu_active(i915))
+		intel_uncore_write(&i915->uncore, vgtif_reg(display_ready),
+				   VGT_DRV_DISPLAY_READY);
+}
+
+bool intel_vgpu_active(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->vgpu.active;
+}
+
 bool intel_vgpu_has_full_ppgtt(struct drm_i915_private *dev_priv)
 {
 	return dev_priv->vgpu.caps & VGT_CAPS_FULL_PPGTT;
 }
 
+bool intel_vgpu_has_hwsp_emulation(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->vgpu.caps & VGT_CAPS_HWSP_EMULATION;
+}
+
+bool intel_vgpu_has_huge_gtt(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->vgpu.caps & VGT_CAPS_HUGE_GTT;
+}
+
 struct _balloon_info_ {
 	/*
 	 * There are up to 2 regions per mappable/unmappable graphic
diff --git a/drivers/gpu/drm/i915/i915_vgpu.h b/drivers/gpu/drm/i915/i915_vgpu.h
index 8b3663dad193..ffbb77d08048 100644
--- a/drivers/gpu/drm/i915/i915_vgpu.h
+++ b/drivers/gpu/drm/i915/i915_vgpu.h
@@ -24,24 +24,17 @@
 #ifndef _I915_VGPU_H_
 #define _I915_VGPU_H_
 
-#include "i915_drv.h"
-#include "i915_pvinfo.h"
+#include <linux/types.h>
 
-void i915_detect_vgpu(struct drm_i915_private *dev_priv);
+struct drm_i915_private;
+struct i915_ggtt;
 
-bool intel_vgpu_has_full_ppgtt(struct drm_i915_private *dev_priv);
-
-static inline bool
-intel_vgpu_has_hwsp_emulation(struct drm_i915_private *dev_priv)
-{
-	return dev_priv->vgpu.caps & VGT_CAPS_HWSP_EMULATION;
-}
-
-static inline bool
-intel_vgpu_has_huge_gtt(struct drm_i915_private *dev_priv)
-{
-	return dev_priv->vgpu.caps & VGT_CAPS_HUGE_GTT;
-}
+void intel_vgpu_detect(struct drm_i915_private *i915);
+bool intel_vgpu_active(struct drm_i915_private *i915);
+void intel_vgpu_register(struct drm_i915_private *i915);
+bool intel_vgpu_has_full_ppgtt(struct drm_i915_private *i915);
+bool intel_vgpu_has_hwsp_emulation(struct drm_i915_private *i915);
+bool intel_vgpu_has_huge_gtt(struct drm_i915_private *i915);
 
 int intel_vgt_balloon(struct i915_ggtt *ggtt);
 void intel_vgt_deballoon(struct i915_ggtt *ggtt);
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index 298ca4316e65..5b3efb43a8ef 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -641,7 +641,6 @@ i915_vma_insert(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
 	u64 start, end;
 	int ret;
 
-	GEM_BUG_ON(i915_vma_is_closed(vma));
 	GEM_BUG_ON(i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND));
 	GEM_BUG_ON(drm_mm_node_allocated(&vma->node));
 
@@ -1174,7 +1173,7 @@ int __i915_vma_move_to_active(struct i915_vma *vma, struct i915_request *rq)
 	GEM_BUG_ON(!i915_vma_is_pinned(vma));
 
 	/* Wait for the vma to be bound before we start! */
-	err = i915_request_await_active(rq, &vma->active);
+	err = i915_request_await_active(rq, &vma->active, 0);
 	if (err)
 		return err;
 
diff --git a/drivers/gpu/drm/i915/i915_vma_types.h b/drivers/gpu/drm/i915/i915_vma_types.h
index e0942efd5236..63831cdb7402 100644
--- a/drivers/gpu/drm/i915/i915_vma_types.h
+++ b/drivers/gpu/drm/i915/i915_vma_types.h
@@ -273,21 +273,10 @@ struct i915_vma {
 	struct rb_node obj_node;
 	struct hlist_node obj_hash;
 
-	/** This vma's place in the execbuf reservation list */
-	struct list_head exec_link;
-	struct list_head reloc_link;
-
 	/** This vma's place in the eviction list */
 	struct list_head evict_link;
 
 	struct list_head closed_link;
-
-	/**
-	 * Used for performing relocations during execbuffer insertion.
-	 */
-	unsigned int *exec_flags;
-	struct hlist_node exec_node;
-	u32 exec_handle;
 };
 
 #endif
diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c
index 8e99ad097830..d7fe12734db8 100644
--- a/drivers/gpu/drm/i915/intel_device_info.c
+++ b/drivers/gpu/drm/i915/intel_device_info.c
@@ -23,6 +23,7 @@
  */
 
 #include <drm/drm_print.h>
+#include <drm/i915_pciids.h>
 
 #include "display/intel_cdclk.h"
 #include "intel_device_info.h"
diff --git a/drivers/gpu/drm/i915/intel_dram.c b/drivers/gpu/drm/i915/intel_dram.c
new file mode 100644
index 000000000000..6b922efb1d7c
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_dram.c
@@ -0,0 +1,500 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2020 Intel Corporation
+ */
+
+#include "i915_drv.h"
+#include "intel_dram.h"
+
+struct dram_dimm_info {
+	u8 size, width, ranks;
+};
+
+struct dram_channel_info {
+	struct dram_dimm_info dimm_l, dimm_s;
+	u8 ranks;
+	bool is_16gb_dimm;
+};
+
+#define DRAM_TYPE_STR(type) [INTEL_DRAM_ ## type] = #type
+
+static const char *intel_dram_type_str(enum intel_dram_type type)
+{
+	static const char * const str[] = {
+		DRAM_TYPE_STR(UNKNOWN),
+		DRAM_TYPE_STR(DDR3),
+		DRAM_TYPE_STR(DDR4),
+		DRAM_TYPE_STR(LPDDR3),
+		DRAM_TYPE_STR(LPDDR4),
+	};
+
+	if (type >= ARRAY_SIZE(str))
+		type = INTEL_DRAM_UNKNOWN;
+
+	return str[type];
+}
+
+#undef DRAM_TYPE_STR
+
+static int intel_dimm_num_devices(const struct dram_dimm_info *dimm)
+{
+	return dimm->ranks * 64 / (dimm->width ?: 1);
+}
+
+/* Returns total GB for the whole DIMM */
+static int skl_get_dimm_size(u16 val)
+{
+	return val & SKL_DRAM_SIZE_MASK;
+}
+
+static int skl_get_dimm_width(u16 val)
+{
+	if (skl_get_dimm_size(val) == 0)
+		return 0;
+
+	switch (val & SKL_DRAM_WIDTH_MASK) {
+	case SKL_DRAM_WIDTH_X8:
+	case SKL_DRAM_WIDTH_X16:
+	case SKL_DRAM_WIDTH_X32:
+		val = (val & SKL_DRAM_WIDTH_MASK) >> SKL_DRAM_WIDTH_SHIFT;
+		return 8 << val;
+	default:
+		MISSING_CASE(val);
+		return 0;
+	}
+}
+
+static int skl_get_dimm_ranks(u16 val)
+{
+	if (skl_get_dimm_size(val) == 0)
+		return 0;
+
+	val = (val & SKL_DRAM_RANK_MASK) >> SKL_DRAM_RANK_SHIFT;
+
+	return val + 1;
+}
+
+/* Returns total GB for the whole DIMM */
+static int cnl_get_dimm_size(u16 val)
+{
+	return (val & CNL_DRAM_SIZE_MASK) / 2;
+}
+
+static int cnl_get_dimm_width(u16 val)
+{
+	if (cnl_get_dimm_size(val) == 0)
+		return 0;
+
+	switch (val & CNL_DRAM_WIDTH_MASK) {
+	case CNL_DRAM_WIDTH_X8:
+	case CNL_DRAM_WIDTH_X16:
+	case CNL_DRAM_WIDTH_X32:
+		val = (val & CNL_DRAM_WIDTH_MASK) >> CNL_DRAM_WIDTH_SHIFT;
+		return 8 << val;
+	default:
+		MISSING_CASE(val);
+		return 0;
+	}
+}
+
+static int cnl_get_dimm_ranks(u16 val)
+{
+	if (cnl_get_dimm_size(val) == 0)
+		return 0;
+
+	val = (val & CNL_DRAM_RANK_MASK) >> CNL_DRAM_RANK_SHIFT;
+
+	return val + 1;
+}
+
+static bool
+skl_is_16gb_dimm(const struct dram_dimm_info *dimm)
+{
+	/* Convert total GB to Gb per DRAM device */
+	return 8 * dimm->size / (intel_dimm_num_devices(dimm) ?: 1) == 16;
+}
+
+static void
+skl_dram_get_dimm_info(struct drm_i915_private *i915,
+		       struct dram_dimm_info *dimm,
+		       int channel, char dimm_name, u16 val)
+{
+	if (INTEL_GEN(i915) >= 10) {
+		dimm->size = cnl_get_dimm_size(val);
+		dimm->width = cnl_get_dimm_width(val);
+		dimm->ranks = cnl_get_dimm_ranks(val);
+	} else {
+		dimm->size = skl_get_dimm_size(val);
+		dimm->width = skl_get_dimm_width(val);
+		dimm->ranks = skl_get_dimm_ranks(val);
+	}
+
+	drm_dbg_kms(&i915->drm,
+		    "CH%u DIMM %c size: %u GB, width: X%u, ranks: %u, 16Gb DIMMs: %s\n",
+		    channel, dimm_name, dimm->size, dimm->width, dimm->ranks,
+		    yesno(skl_is_16gb_dimm(dimm)));
+}
+
+static int
+skl_dram_get_channel_info(struct drm_i915_private *i915,
+			  struct dram_channel_info *ch,
+			  int channel, u32 val)
+{
+	skl_dram_get_dimm_info(i915, &ch->dimm_l,
+			       channel, 'L', val & 0xffff);
+	skl_dram_get_dimm_info(i915, &ch->dimm_s,
+			       channel, 'S', val >> 16);
+
+	if (ch->dimm_l.size == 0 && ch->dimm_s.size == 0) {
+		drm_dbg_kms(&i915->drm, "CH%u not populated\n", channel);
+		return -EINVAL;
+	}
+
+	if (ch->dimm_l.ranks == 2 || ch->dimm_s.ranks == 2)
+		ch->ranks = 2;
+	else if (ch->dimm_l.ranks == 1 && ch->dimm_s.ranks == 1)
+		ch->ranks = 2;
+	else
+		ch->ranks = 1;
+
+	ch->is_16gb_dimm = skl_is_16gb_dimm(&ch->dimm_l) ||
+		skl_is_16gb_dimm(&ch->dimm_s);
+
+	drm_dbg_kms(&i915->drm, "CH%u ranks: %u, 16Gb DIMMs: %s\n",
+		    channel, ch->ranks, yesno(ch->is_16gb_dimm));
+
+	return 0;
+}
+
+static bool
+intel_is_dram_symmetric(const struct dram_channel_info *ch0,
+			const struct dram_channel_info *ch1)
+{
+	return !memcmp(ch0, ch1, sizeof(*ch0)) &&
+		(ch0->dimm_s.size == 0 ||
+		 !memcmp(&ch0->dimm_l, &ch0->dimm_s, sizeof(ch0->dimm_l)));
+}
+
+static int
+skl_dram_get_channels_info(struct drm_i915_private *i915)
+{
+	struct dram_info *dram_info = &i915->dram_info;
+	struct dram_channel_info ch0 = {}, ch1 = {};
+	u32 val;
+	int ret;
+
+	val = intel_uncore_read(&i915->uncore,
+				SKL_MAD_DIMM_CH0_0_0_0_MCHBAR_MCMAIN);
+	ret = skl_dram_get_channel_info(i915, &ch0, 0, val);
+	if (ret == 0)
+		dram_info->num_channels++;
+
+	val = intel_uncore_read(&i915->uncore,
+				SKL_MAD_DIMM_CH1_0_0_0_MCHBAR_MCMAIN);
+	ret = skl_dram_get_channel_info(i915, &ch1, 1, val);
+	if (ret == 0)
+		dram_info->num_channels++;
+
+	if (dram_info->num_channels == 0) {
+		drm_info(&i915->drm, "Number of memory channels is zero\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * If any of the channel is single rank channel, worst case output
+	 * will be same as if single rank memory, so consider single rank
+	 * memory.
+	 */
+	if (ch0.ranks == 1 || ch1.ranks == 1)
+		dram_info->ranks = 1;
+	else
+		dram_info->ranks = max(ch0.ranks, ch1.ranks);
+
+	if (dram_info->ranks == 0) {
+		drm_info(&i915->drm, "couldn't get memory rank information\n");
+		return -EINVAL;
+	}
+
+	dram_info->is_16gb_dimm = ch0.is_16gb_dimm || ch1.is_16gb_dimm;
+
+	dram_info->symmetric_memory = intel_is_dram_symmetric(&ch0, &ch1);
+
+	drm_dbg_kms(&i915->drm, "Memory configuration is symmetric? %s\n",
+		    yesno(dram_info->symmetric_memory));
+
+	return 0;
+}
+
+static enum intel_dram_type
+skl_get_dram_type(struct drm_i915_private *i915)
+{
+	u32 val;
+
+	val = intel_uncore_read(&i915->uncore,
+				SKL_MAD_INTER_CHANNEL_0_0_0_MCHBAR_MCMAIN);
+
+	switch (val & SKL_DRAM_DDR_TYPE_MASK) {
+	case SKL_DRAM_DDR_TYPE_DDR3:
+		return INTEL_DRAM_DDR3;
+	case SKL_DRAM_DDR_TYPE_DDR4:
+		return INTEL_DRAM_DDR4;
+	case SKL_DRAM_DDR_TYPE_LPDDR3:
+		return INTEL_DRAM_LPDDR3;
+	case SKL_DRAM_DDR_TYPE_LPDDR4:
+		return INTEL_DRAM_LPDDR4;
+	default:
+		MISSING_CASE(val);
+		return INTEL_DRAM_UNKNOWN;
+	}
+}
+
+static int
+skl_get_dram_info(struct drm_i915_private *i915)
+{
+	struct dram_info *dram_info = &i915->dram_info;
+	u32 mem_freq_khz, val;
+	int ret;
+
+	dram_info->type = skl_get_dram_type(i915);
+	drm_dbg_kms(&i915->drm, "DRAM type: %s\n",
+		    intel_dram_type_str(dram_info->type));
+
+	ret = skl_dram_get_channels_info(i915);
+	if (ret)
+		return ret;
+
+	val = intel_uncore_read(&i915->uncore,
+				SKL_MC_BIOS_DATA_0_0_0_MCHBAR_PCU);
+	mem_freq_khz = DIV_ROUND_UP((val & SKL_REQ_DATA_MASK) *
+				    SKL_MEMORY_FREQ_MULTIPLIER_HZ, 1000);
+
+	dram_info->bandwidth_kbps = dram_info->num_channels *
+		mem_freq_khz * 8;
+
+	if (dram_info->bandwidth_kbps == 0) {
+		drm_info(&i915->drm,
+			 "Couldn't get system memory bandwidth\n");
+		return -EINVAL;
+	}
+
+	dram_info->valid = true;
+	return 0;
+}
+
+/* Returns Gb per DRAM device */
+static int bxt_get_dimm_size(u32 val)
+{
+	switch (val & BXT_DRAM_SIZE_MASK) {
+	case BXT_DRAM_SIZE_4GBIT:
+		return 4;
+	case BXT_DRAM_SIZE_6GBIT:
+		return 6;
+	case BXT_DRAM_SIZE_8GBIT:
+		return 8;
+	case BXT_DRAM_SIZE_12GBIT:
+		return 12;
+	case BXT_DRAM_SIZE_16GBIT:
+		return 16;
+	default:
+		MISSING_CASE(val);
+		return 0;
+	}
+}
+
+static int bxt_get_dimm_width(u32 val)
+{
+	if (!bxt_get_dimm_size(val))
+		return 0;
+
+	val = (val & BXT_DRAM_WIDTH_MASK) >> BXT_DRAM_WIDTH_SHIFT;
+
+	return 8 << val;
+}
+
+static int bxt_get_dimm_ranks(u32 val)
+{
+	if (!bxt_get_dimm_size(val))
+		return 0;
+
+	switch (val & BXT_DRAM_RANK_MASK) {
+	case BXT_DRAM_RANK_SINGLE:
+		return 1;
+	case BXT_DRAM_RANK_DUAL:
+		return 2;
+	default:
+		MISSING_CASE(val);
+		return 0;
+	}
+}
+
+static enum intel_dram_type bxt_get_dimm_type(u32 val)
+{
+	if (!bxt_get_dimm_size(val))
+		return INTEL_DRAM_UNKNOWN;
+
+	switch (val & BXT_DRAM_TYPE_MASK) {
+	case BXT_DRAM_TYPE_DDR3:
+		return INTEL_DRAM_DDR3;
+	case BXT_DRAM_TYPE_LPDDR3:
+		return INTEL_DRAM_LPDDR3;
+	case BXT_DRAM_TYPE_DDR4:
+		return INTEL_DRAM_DDR4;
+	case BXT_DRAM_TYPE_LPDDR4:
+		return INTEL_DRAM_LPDDR4;
+	default:
+		MISSING_CASE(val);
+		return INTEL_DRAM_UNKNOWN;
+	}
+}
+
+static void bxt_get_dimm_info(struct dram_dimm_info *dimm, u32 val)
+{
+	dimm->width = bxt_get_dimm_width(val);
+	dimm->ranks = bxt_get_dimm_ranks(val);
+
+	/*
+	 * Size in register is Gb per DRAM device. Convert to total
+	 * GB to match the way we report this for non-LP platforms.
+	 */
+	dimm->size = bxt_get_dimm_size(val) * intel_dimm_num_devices(dimm) / 8;
+}
+
+static int bxt_get_dram_info(struct drm_i915_private *i915)
+{
+	struct dram_info *dram_info = &i915->dram_info;
+	u32 dram_channels;
+	u32 mem_freq_khz, val;
+	u8 num_active_channels;
+	int i;
+
+	val = intel_uncore_read(&i915->uncore, BXT_P_CR_MC_BIOS_REQ_0_0_0);
+	mem_freq_khz = DIV_ROUND_UP((val & BXT_REQ_DATA_MASK) *
+				    BXT_MEMORY_FREQ_MULTIPLIER_HZ, 1000);
+
+	dram_channels = val & BXT_DRAM_CHANNEL_ACTIVE_MASK;
+	num_active_channels = hweight32(dram_channels);
+
+	/* Each active bit represents 4-byte channel */
+	dram_info->bandwidth_kbps = (mem_freq_khz * num_active_channels * 4);
+
+	if (dram_info->bandwidth_kbps == 0) {
+		drm_info(&i915->drm,
+			 "Couldn't get system memory bandwidth\n");
+		return -EINVAL;
+	}
+
+	/*
+	 * Now read each DUNIT8/9/10/11 to check the rank of each dimms.
+	 */
+	for (i = BXT_D_CR_DRP0_DUNIT_START; i <= BXT_D_CR_DRP0_DUNIT_END; i++) {
+		struct dram_dimm_info dimm;
+		enum intel_dram_type type;
+
+		val = intel_uncore_read(&i915->uncore, BXT_D_CR_DRP0_DUNIT(i));
+		if (val == 0xFFFFFFFF)
+			continue;
+
+		dram_info->num_channels++;
+
+		bxt_get_dimm_info(&dimm, val);
+		type = bxt_get_dimm_type(val);
+
+		drm_WARN_ON(&i915->drm, type != INTEL_DRAM_UNKNOWN &&
+			    dram_info->type != INTEL_DRAM_UNKNOWN &&
+			    dram_info->type != type);
+
+		drm_dbg_kms(&i915->drm,
+			    "CH%u DIMM size: %u GB, width: X%u, ranks: %u, type: %s\n",
+			    i - BXT_D_CR_DRP0_DUNIT_START,
+			    dimm.size, dimm.width, dimm.ranks,
+			    intel_dram_type_str(type));
+
+		/*
+		 * If any of the channel is single rank channel,
+		 * worst case output will be same as if single rank
+		 * memory, so consider single rank memory.
+		 */
+		if (dram_info->ranks == 0)
+			dram_info->ranks = dimm.ranks;
+		else if (dimm.ranks == 1)
+			dram_info->ranks = 1;
+
+		if (type != INTEL_DRAM_UNKNOWN)
+			dram_info->type = type;
+	}
+
+	if (dram_info->type == INTEL_DRAM_UNKNOWN || dram_info->ranks == 0) {
+		drm_info(&i915->drm, "couldn't get memory information\n");
+		return -EINVAL;
+	}
+
+	dram_info->valid = true;
+
+	return 0;
+}
+
+void intel_dram_detect(struct drm_i915_private *i915)
+{
+	struct dram_info *dram_info = &i915->dram_info;
+	int ret;
+
+	/*
+	 * Assume 16Gb DIMMs are present until proven otherwise.
+	 * This is only used for the level 0 watermark latency
+	 * w/a which does not apply to bxt/glk.
+	 */
+	dram_info->is_16gb_dimm = !IS_GEN9_LP(i915);
+
+	if (INTEL_GEN(i915) < 9 || !HAS_DISPLAY(i915))
+		return;
+
+	if (IS_GEN9_LP(i915))
+		ret = bxt_get_dram_info(i915);
+	else
+		ret = skl_get_dram_info(i915);
+	if (ret)
+		return;
+
+	drm_dbg_kms(&i915->drm, "DRAM bandwidth: %u kBps, channels: %u\n",
+		    dram_info->bandwidth_kbps, dram_info->num_channels);
+
+	drm_dbg_kms(&i915->drm, "DRAM ranks: %u, 16Gb DIMMs: %s\n",
+		    dram_info->ranks, yesno(dram_info->is_16gb_dimm));
+}
+
+static u32 gen9_edram_size_mb(struct drm_i915_private *i915, u32 cap)
+{
+	static const u8 ways[8] = { 4, 8, 12, 16, 16, 16, 16, 16 };
+	static const u8 sets[4] = { 1, 1, 2, 2 };
+
+	return EDRAM_NUM_BANKS(cap) *
+		ways[EDRAM_WAYS_IDX(cap)] *
+		sets[EDRAM_SETS_IDX(cap)];
+}
+
+void intel_dram_edram_detect(struct drm_i915_private *i915)
+{
+	u32 edram_cap = 0;
+
+	if (!(IS_HASWELL(i915) || IS_BROADWELL(i915) || INTEL_GEN(i915) >= 9))
+		return;
+
+	edram_cap = __raw_uncore_read32(&i915->uncore, HSW_EDRAM_CAP);
+
+	/* NB: We can't write IDICR yet because we don't have gt funcs set up */
+
+	if (!(edram_cap & EDRAM_ENABLED))
+		return;
+
+	/*
+	 * The needed capability bits for size calculation are not there with
+	 * pre gen9 so return 128MB always.
+	 */
+	if (INTEL_GEN(i915) < 9)
+		i915->edram_size_mb = 128;
+	else
+		i915->edram_size_mb = gen9_edram_size_mb(i915, edram_cap);
+
+	dev_info(i915->drm.dev,
+		 "Found %uMB of eDRAM\n", i915->edram_size_mb);
+}
diff --git a/drivers/gpu/drm/i915/intel_dram.h b/drivers/gpu/drm/i915/intel_dram.h
new file mode 100644
index 000000000000..4ba13c13162c
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_dram.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2020 Intel Corporation
+ */
+
+#ifndef __INTEL_DRAM_H__
+#define __INTEL_DRAM_H__
+
+struct drm_i915_private;
+
+void intel_dram_edram_detect(struct drm_i915_private *i915);
+void intel_dram_detect(struct drm_i915_private *i915);
+
+#endif /* __INTEL_DRAM_H__ */
diff --git a/drivers/gpu/drm/i915/intel_gvt.c b/drivers/gpu/drm/i915/intel_gvt.c
index e73fd752adef..21b91313cc5d 100644
--- a/drivers/gpu/drm/i915/intel_gvt.c
+++ b/drivers/gpu/drm/i915/intel_gvt.c
@@ -22,6 +22,7 @@
  */
 
 #include "i915_drv.h"
+#include "i915_vgpu.h"
 #include "intel_gvt.h"
 
 /**
@@ -124,6 +125,11 @@ bail:
 	return 0;
 }
 
+static inline bool intel_gvt_active(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->gvt;
+}
+
 /**
  * intel_gvt_driver_remove - cleanup GVT components when i915 driver is
  *			     unbinding
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 22aa205793e5..8375054ba27d 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -40,12 +40,36 @@
 #include "gt/intel_llc.h"
 
 #include "i915_drv.h"
+#include "i915_fixed.h"
 #include "i915_irq.h"
 #include "i915_trace.h"
 #include "intel_pm.h"
 #include "intel_sideband.h"
 #include "../../../platform/x86/intel_ips.h"
 
+/* Stores plane specific WM parameters */
+struct skl_wm_params {
+	bool x_tiled, y_tiled;
+	bool rc_surface;
+	bool is_planar;
+	u32 width;
+	u8 cpp;
+	u32 plane_pixel_rate;
+	u32 y_min_scanlines;
+	u32 plane_bytes_per_line;
+	uint_fixed_16_16_t plane_blocks_per_line;
+	uint_fixed_16_16_t y_tile_minimum;
+	u32 linetime_us;
+	u32 dbuf_block_size;
+};
+
+/* used in computing the new watermarks state */
+struct intel_wm_config {
+	unsigned int num_pipes_active;
+	bool sprites_enabled;
+	bool sprites_scaled;
+};
+
 static void gen9_init_clock_gating(struct drm_i915_private *dev_priv)
 {
 	if (HAS_LLC(dev_priv)) {
@@ -128,16 +152,6 @@ static void glk_init_clock_gating(struct drm_i915_private *dev_priv)
 	 */
 	I915_WRITE(GEN9_CLKGATE_DIS_0, I915_READ(GEN9_CLKGATE_DIS_0) |
 		   PWM1_GATING_DIS | PWM2_GATING_DIS);
-
-	/* WaDDIIOTimeout:glk */
-	if (IS_GLK_REVID(dev_priv, 0, GLK_REVID_A1)) {
-		u32 val = I915_READ(CHICKEN_MISC_2);
-		val &= ~(GLK_CL0_PWR_DOWN |
-			 GLK_CL1_PWR_DOWN |
-			 GLK_CL2_PWR_DOWN);
-		I915_WRITE(CHICKEN_MISC_2, val);
-	}
-
 }
 
 static void pnv_get_mem_freq(struct drm_i915_private *dev_priv)
@@ -2776,7 +2790,7 @@ static bool ilk_validate_wm_level(int level,
 }
 
 static void ilk_compute_wm_level(const struct drm_i915_private *dev_priv,
-				 const struct intel_crtc *intel_crtc,
+				 const struct intel_crtc *crtc,
 				 int level,
 				 struct intel_crtc_state *crtc_state,
 				 const struct intel_plane_state *pristate,
@@ -3107,7 +3121,7 @@ static bool ilk_validate_pipe_wm(const struct drm_i915_private *dev_priv,
 static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 {
 	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_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_pipe_wm *pipe_wm;
 	struct intel_plane *plane;
 	const struct intel_plane_state *plane_state;
@@ -3147,7 +3161,7 @@ static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 		usable_level = 0;
 
 	memset(&pipe_wm->wm, 0, sizeof(pipe_wm->wm));
-	ilk_compute_wm_level(dev_priv, intel_crtc, 0, crtc_state,
+	ilk_compute_wm_level(dev_priv, crtc, 0, crtc_state,
 			     pristate, sprstate, curstate, &pipe_wm->wm[0]);
 
 	if (!ilk_validate_pipe_wm(dev_priv, pipe_wm))
@@ -3158,7 +3172,7 @@ static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	for (level = 1; level <= usable_level; level++) {
 		struct intel_wm_level *wm = &pipe_wm->wm[level];
 
-		ilk_compute_wm_level(dev_priv, intel_crtc, level, crtc_state,
+		ilk_compute_wm_level(dev_priv, crtc, level, crtc_state,
 				     pristate, sprstate, curstate, wm);
 
 		/*
@@ -3843,7 +3857,7 @@ static u16 intel_get_ddb_size(struct drm_i915_private *dev_priv)
 }
 
 static u8 skl_compute_dbuf_slices(const struct intel_crtc_state *crtc_state,
-				  u32 active_pipes);
+				  u8 active_pipes);
 
 static void
 skl_ddb_get_pipe_allocation_limits(struct drm_i915_private *dev_priv,
@@ -4184,50 +4198,51 @@ static const struct dbuf_slice_conf_entry icl_allowed_dbufs[] =
 	{
 		.active_pipes = BIT(PIPE_A),
 		.dbuf_mask = {
-			[PIPE_A] = BIT(DBUF_S1)
-		}
+			[PIPE_A] = BIT(DBUF_S1),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_B),
 		.dbuf_mask = {
-			[PIPE_B] = BIT(DBUF_S1)
-		}
+			[PIPE_B] = BIT(DBUF_S1),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_B),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
-			[PIPE_B] = BIT(DBUF_S2)
-		}
+			[PIPE_B] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_C),
 		.dbuf_mask = {
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_C),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_B) | BIT(PIPE_C),
 		.dbuf_mask = {
 			[PIPE_B] = BIT(DBUF_S1),
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
 			[PIPE_B] = BIT(DBUF_S1),
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
+	{}
 };
 
 /*
@@ -4246,100 +4261,100 @@ static const struct dbuf_slice_conf_entry tgl_allowed_dbufs[] =
 	{
 		.active_pipes = BIT(PIPE_A),
 		.dbuf_mask = {
-			[PIPE_A] = BIT(DBUF_S1) | BIT(DBUF_S2)
-		}
+			[PIPE_A] = BIT(DBUF_S1) | BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_B),
 		.dbuf_mask = {
-			[PIPE_B] = BIT(DBUF_S1) | BIT(DBUF_S2)
-		}
+			[PIPE_B] = BIT(DBUF_S1) | BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_B),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S2),
-			[PIPE_B] = BIT(DBUF_S1)
-		}
+			[PIPE_B] = BIT(DBUF_S1),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_C),
 		.dbuf_mask = {
-			[PIPE_C] = BIT(DBUF_S2) | BIT(DBUF_S1)
-		}
+			[PIPE_C] = BIT(DBUF_S2) | BIT(DBUF_S1),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_C),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_B) | BIT(PIPE_C),
 		.dbuf_mask = {
 			[PIPE_B] = BIT(DBUF_S1),
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
 			[PIPE_B] = BIT(DBUF_S1),
-			[PIPE_C] = BIT(DBUF_S2)
-		}
+			[PIPE_C] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_D),
 		.dbuf_mask = {
-			[PIPE_D] = BIT(DBUF_S2) | BIT(DBUF_S1)
-		}
+			[PIPE_D] = BIT(DBUF_S2) | BIT(DBUF_S1),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_D),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_B) | BIT(PIPE_D),
 		.dbuf_mask = {
 			[PIPE_B] = BIT(DBUF_S1),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_D),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
 			[PIPE_B] = BIT(DBUF_S1),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_C) | BIT(PIPE_D),
 		.dbuf_mask = {
 			[PIPE_C] = BIT(DBUF_S1),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_C) | BIT(PIPE_D),
 		.dbuf_mask = {
 			[PIPE_A] = BIT(DBUF_S1),
 			[PIPE_C] = BIT(DBUF_S2),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_B) | BIT(PIPE_C) | BIT(PIPE_D),
 		.dbuf_mask = {
 			[PIPE_B] = BIT(DBUF_S1),
 			[PIPE_C] = BIT(DBUF_S2),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
 	{
 		.active_pipes = BIT(PIPE_A) | BIT(PIPE_B) | BIT(PIPE_C) | BIT(PIPE_D),
@@ -4347,19 +4362,18 @@ static const struct dbuf_slice_conf_entry tgl_allowed_dbufs[] =
 			[PIPE_A] = BIT(DBUF_S1),
 			[PIPE_B] = BIT(DBUF_S1),
 			[PIPE_C] = BIT(DBUF_S2),
-			[PIPE_D] = BIT(DBUF_S2)
-		}
+			[PIPE_D] = BIT(DBUF_S2),
+		},
 	},
+	{}
 };
 
-static u8 compute_dbuf_slices(enum pipe pipe,
-			      u32 active_pipes,
-			      const struct dbuf_slice_conf_entry *dbuf_slices,
-			      int size)
+static u8 compute_dbuf_slices(enum pipe pipe, u8 active_pipes,
+			      const struct dbuf_slice_conf_entry *dbuf_slices)
 {
 	int i;
 
-	for (i = 0; i < size; i++) {
+	for (i = 0; i < dbuf_slices[i].active_pipes; i++) {
 		if (dbuf_slices[i].active_pipes == active_pipes)
 			return dbuf_slices[i].dbuf_mask[pipe];
 	}
@@ -4371,8 +4385,7 @@ static u8 compute_dbuf_slices(enum pipe pipe,
  * returns correspondent DBuf slice mask as stated in BSpec for particular
  * platform.
  */
-static u32 icl_compute_dbuf_slices(enum pipe pipe,
-				   u32 active_pipes)
+static u8 icl_compute_dbuf_slices(enum pipe pipe, u8 active_pipes)
 {
 	/*
 	 * FIXME: For ICL this is still a bit unclear as prev BSpec revision
@@ -4386,32 +4399,25 @@ static u32 icl_compute_dbuf_slices(enum pipe pipe,
 	 * still here - we will need it once those additional constraints
 	 * pop up.
 	 */
-	return compute_dbuf_slices(pipe, active_pipes,
-				   icl_allowed_dbufs,
-				   ARRAY_SIZE(icl_allowed_dbufs));
+	return compute_dbuf_slices(pipe, active_pipes, icl_allowed_dbufs);
 }
 
-static u32 tgl_compute_dbuf_slices(enum pipe pipe,
-				   u32 active_pipes)
+static u8 tgl_compute_dbuf_slices(enum pipe pipe, u8 active_pipes)
 {
-	return compute_dbuf_slices(pipe, active_pipes,
-				   tgl_allowed_dbufs,
-				   ARRAY_SIZE(tgl_allowed_dbufs));
+	return compute_dbuf_slices(pipe, active_pipes, tgl_allowed_dbufs);
 }
 
 static u8 skl_compute_dbuf_slices(const struct intel_crtc_state *crtc_state,
-				  u32 active_pipes)
+				  u8 active_pipes)
 {
 	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 (IS_GEN(dev_priv, 12))
-		return tgl_compute_dbuf_slices(pipe,
-					       active_pipes);
+		return tgl_compute_dbuf_slices(pipe, active_pipes);
 	else if (IS_GEN(dev_priv, 11))
-		return icl_compute_dbuf_slices(pipe,
-					       active_pipes);
+		return icl_compute_dbuf_slices(pipe, active_pipes);
 	/*
 	 * For anything else just return one slice yet.
 	 * Should be extended for other platforms.
@@ -4470,14 +4476,10 @@ 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->uapi.state;
 	struct intel_plane *plane;
 	const struct intel_plane_state *plane_state;
 	u64 total_data_rate = 0;
 
-	if (WARN_ON(!state))
-		return 0;
-
 	/* Calculate and cache data rate for each plane */
 	intel_atomic_crtc_state_for_each_plane_state(plane, plane_state, crtc_state) {
 		enum plane_id plane_id = plane->id;
@@ -4505,9 +4507,6 @@ 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->uapi.state))
-		return 0;
-
 	/* Calculate and cache data rate for each plane */
 	intel_atomic_crtc_state_for_each_plane_state(plane, plane_state, crtc_state) {
 		enum plane_id plane_id = plane->id;
@@ -4548,10 +4547,8 @@ icl_get_total_relative_data_rate(struct intel_crtc_state *crtc_state,
 static int
 skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 {
-	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 intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct skl_ddb_entry *alloc = &crtc_state->wm.skl.ddb;
 	u16 alloc_size, start = 0;
 	u16 total[I915_MAX_PLANES] = {};
@@ -4568,9 +4565,6 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 	memset(crtc_state->wm.skl.plane_ddb_y, 0, sizeof(crtc_state->wm.skl.plane_ddb_y));
 	memset(crtc_state->wm.skl.plane_ddb_uv, 0, sizeof(crtc_state->wm.skl.plane_ddb_uv));
 
-	if (drm_WARN_ON(&dev_priv->drm, !state))
-		return 0;
-
 	if (!crtc_state->hw.active) {
 		alloc->start = alloc->end = 0;
 		return 0;
@@ -4609,7 +4603,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 	 */
 	for (level = ilk_wm_max_level(dev_priv); level >= 0; level--) {
 		blocks = 0;
-		for_each_plane_id_on_crtc(intel_crtc, plane_id) {
+		for_each_plane_id_on_crtc(crtc, plane_id) {
 			const struct skl_plane_wm *wm =
 				&crtc_state->wm.skl.optimal.planes[plane_id];
 
@@ -4646,7 +4640,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 	 * watermark level, plus an extra share of the leftover blocks
 	 * proportional to its relative data rate.
 	 */
-	for_each_plane_id_on_crtc(intel_crtc, plane_id) {
+	for_each_plane_id_on_crtc(crtc, plane_id) {
 		const struct skl_plane_wm *wm =
 			&crtc_state->wm.skl.optimal.planes[plane_id];
 		u64 rate;
@@ -4685,7 +4679,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 
 	/* Set the actual DDB start/end points for each plane */
 	start = alloc->start;
-	for_each_plane_id_on_crtc(intel_crtc, plane_id) {
+	for_each_plane_id_on_crtc(crtc, plane_id) {
 		struct skl_ddb_entry *plane_alloc =
 			&crtc_state->wm.skl.plane_ddb_y[plane_id];
 		struct skl_ddb_entry *uv_plane_alloc =
@@ -4719,7 +4713,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 	 * that aren't actually possible.
 	 */
 	for (level++; level <= ilk_wm_max_level(dev_priv); level++) {
-		for_each_plane_id_on_crtc(intel_crtc, plane_id) {
+		for_each_plane_id_on_crtc(crtc, plane_id) {
 			struct skl_plane_wm *wm =
 				&crtc_state->wm.skl.optimal.planes[plane_id];
 
@@ -4756,7 +4750,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state)
 	 * Go back and disable the transition watermark if it turns out we
 	 * don't have enough DDB blocks for it.
 	 */
-	for_each_plane_id_on_crtc(intel_crtc, plane_id) {
+	for_each_plane_id_on_crtc(crtc, plane_id) {
 		struct skl_plane_wm *wm =
 			&crtc_state->wm.skl.optimal.planes[plane_id];
 
@@ -5126,21 +5120,30 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *crtc_state,
 {
 	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 */
+	u16 trans_min, trans_amount, trans_y_tile_min;
 	u16 wm0_sel_res_b, trans_offset_b, res_blocks;
 
-	/* Transition WM are not recommended by HW team for GEN9 */
-	if (INTEL_GEN(dev_priv) <= 9)
-		return;
-
 	/* Transition WM don't make any sense if ipc is disabled */
 	if (!dev_priv->ipc_enabled)
 		return;
 
-	trans_min = 14;
+	/*
+	 * WaDisableTWM:skl,kbl,cfl,bxt
+	 * Transition WM are not recommended by HW team for GEN9
+	 */
+	if (IS_GEN9_BC(dev_priv) || IS_BROXTON(dev_priv))
+		return;
+
 	if (INTEL_GEN(dev_priv) >= 11)
 		trans_min = 4;
+	else
+		trans_min = 14;
+
+	/* Display WA #1140: glk,cnl */
+	if (IS_CANNONLAKE(dev_priv) || IS_GEMINILAKE(dev_priv))
+		trans_amount = 0;
+	else
+		trans_amount = 10; /* This is configurable amount */
 
 	trans_offset_b = trans_min + trans_amount;
 
@@ -5167,7 +5170,6 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *crtc_state,
 		/* WA BUG:1938466 add one block for non y-tile planes */
 		if (IS_CNL_REVID(dev_priv, CNL_REVID_A0, CNL_REVID_A0))
 			res_blocks += 1;
-
 	}
 
 	/*
@@ -5410,8 +5412,12 @@ static bool skl_plane_wm_equals(struct drm_i915_private *dev_priv,
 	int level, max_level = ilk_wm_max_level(dev_priv);
 
 	for (level = 0; level <= max_level; level++) {
-		if (!skl_wm_level_equals(&wm1->wm[level], &wm2->wm[level]) ||
-		    !skl_wm_level_equals(&wm1->uv_wm[level], &wm2->uv_wm[level]))
+		/*
+		 * We don't check uv_wm as the hardware doesn't actually
+		 * use it. It only gets used for calculating the required
+		 * ddb allocation.
+		 */
+		if (!skl_wm_level_equals(&wm1->wm[level], &wm2->wm[level]))
 			return false;
 	}
 
@@ -5768,16 +5774,24 @@ skl_compute_wm(struct intel_atomic_state *state)
 		ret = skl_build_pipe_wm(new_crtc_state);
 		if (ret)
 			return ret;
-
-		ret = skl_wm_add_affected_planes(state, crtc);
-		if (ret)
-			return ret;
 	}
 
 	ret = skl_compute_ddb(state);
 	if (ret)
 		return ret;
 
+	/*
+	 * skl_compute_ddb() will have adjusted the final watermarks
+	 * based on how much ddb is available. Now we can actually
+	 * check if the final watermarks changed.
+	 */
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+					    new_crtc_state, i) {
+		ret = skl_wm_add_affected_planes(state, crtc);
+		if (ret)
+			return ret;
+	}
+
 	skl_print_wm_changes(state);
 
 	return 0;
@@ -6812,21 +6826,6 @@ static void icl_init_clock_gating(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN10_DFR_RATIO_EN_AND_CHICKEN,
 		   I915_READ(GEN10_DFR_RATIO_EN_AND_CHICKEN) & ~DFR_DISABLE);
 
-	/* WaEnable32PlaneMode:icl */
-	I915_WRITE(GEN9_CSFE_CHICKEN1_RCS,
-		   _MASKED_BIT_ENABLE(GEN11_ENABLE_32_PLANE_MODE));
-
-	/*
-	 * Wa_1408615072:icl,ehl  (vsunit)
-	 * Wa_1407596294:icl,ehl  (hsunit)
-	 */
-	intel_uncore_rmw(&dev_priv->uncore, UNSLICE_UNIT_LEVEL_CLKGATE,
-			 0, VSUNIT_CLKGATE_DIS | HSUNIT_CLKGATE_DIS);
-
-	/* Wa_1407352427:icl,ehl */
-	intel_uncore_rmw(&dev_priv->uncore, UNSLICE_UNIT_LEVEL_CLKGATE2,
-			 0, PSDUNIT_CLKGATE_DIS);
-
 	/*Wa_14010594013:icl, ehl */
 	intel_uncore_rmw(&dev_priv->uncore, GEN8_CHICKEN_DCPR_1,
 			 0, CNL_DELAY_PMRSP);
@@ -6837,10 +6836,6 @@ static void tgl_init_clock_gating(struct drm_i915_private *dev_priv)
 	u32 vd_pg_enable = 0;
 	unsigned int i;
 
-	/* Wa_1408615072:tgl */
-	intel_uncore_rmw(&dev_priv->uncore, UNSLICE_UNIT_LEVEL_CLKGATE2,
-			 0, VSUNIT_CLKGATE_DIS_TGL);
-
 	/* This is not a WA. Enable VD HCP & MFX_ENC powergate */
 	for (i = 0; i < I915_MAX_VCS; i++) {
 		if (HAS_ENGINE(dev_priv, _VCS(i)))
diff --git a/drivers/gpu/drm/i915/selftests/i915_active.c b/drivers/gpu/drm/i915/selftests/i915_active.c
index ef572a0c2566..68bbb1580162 100644
--- a/drivers/gpu/drm/i915/selftests/i915_active.c
+++ b/drivers/gpu/drm/i915/selftests/i915_active.c
@@ -201,11 +201,57 @@ static int live_active_retire(void *arg)
 	return err;
 }
 
+static int live_active_barrier(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct intel_engine_cs *engine;
+	struct live_active *active;
+	int err = 0;
+
+	/* Check that we get a callback when requests retire upon waiting */
+
+	active = __live_alloc(i915);
+	if (!active)
+		return -ENOMEM;
+
+	err = i915_active_acquire(&active->base);
+	if (err)
+		goto out;
+
+	for_each_uabi_engine(engine, i915) {
+		err = i915_active_acquire_preallocate_barrier(&active->base,
+							      engine);
+		if (err)
+			break;
+
+		i915_active_acquire_barrier(&active->base);
+	}
+
+	i915_active_release(&active->base);
+
+	if (err == 0)
+		err = i915_active_wait(&active->base);
+
+	if (err == 0 && !READ_ONCE(active->retired)) {
+		pr_err("i915_active not retired after flushing barriers!\n");
+		err = -EINVAL;
+	}
+
+out:
+	__live_put(active);
+
+	if (igt_flush_test(i915))
+		err = -EIO;
+
+	return err;
+}
+
 int i915_active_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(live_active_wait),
 		SUBTEST(live_active_retire),
+		SUBTEST(live_active_barrier),
 	};
 
 	if (intel_gt_is_wedged(&i915->gt))
@@ -265,28 +311,40 @@ static void spin_unlock_wait(spinlock_t *lock)
 	spin_unlock_irq(lock);
 }
 
+static void active_flush(struct i915_active *ref,
+			 struct i915_active_fence *active)
+{
+	struct dma_fence *fence;
+
+	fence = xchg(__active_fence_slot(active), NULL);
+	if (!fence)
+		return;
+
+	spin_lock_irq(fence->lock);
+	__list_del_entry(&active->cb.node);
+	spin_unlock_irq(fence->lock); /* serialise with fence->cb_list */
+	atomic_dec(&ref->count);
+
+	GEM_BUG_ON(!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags));
+}
+
 void i915_active_unlock_wait(struct i915_active *ref)
 {
 	if (i915_active_acquire_if_busy(ref)) {
 		struct active_node *it, *n;
 
+		/* Wait for all active callbacks */
 		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);
-		}
+		active_flush(ref, &ref->excl);
+		rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node)
+			active_flush(ref, &it->base);
 		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);
+	spin_unlock_wait(&ref->tree_lock);
 
 	/* ... which may have been on a thread instead */
 	flush_work(&ref->work);
diff --git a/drivers/gpu/drm/i915/selftests/i915_buddy.c b/drivers/gpu/drm/i915/selftests/i915_buddy.c
index 1b856bae67b5..939a6caebb03 100644
--- a/drivers/gpu/drm/i915/selftests/i915_buddy.c
+++ b/drivers/gpu/drm/i915/selftests/i915_buddy.c
@@ -298,10 +298,12 @@ static void igt_mm_config(u64 *size, u64 *chunk_size)
 static int igt_buddy_alloc_smoke(void *arg)
 {
 	struct i915_buddy_mm mm;
-	int max_order;
+	IGT_TIMEOUT(end_time);
+	I915_RND_STATE(prng);
 	u64 chunk_size;
 	u64 mm_size;
-	int err;
+	int *order;
+	int err, i;
 
 	igt_mm_config(&mm_size, &chunk_size);
 
@@ -313,10 +315,16 @@ static int igt_buddy_alloc_smoke(void *arg)
 		return err;
 	}
 
-	for (max_order = mm.max_order; max_order >= 0; max_order--) {
+	order = i915_random_order(mm.max_order + 1, &prng);
+	if (!order)
+		goto out_fini;
+
+	for (i = 0; i <= mm.max_order; ++i) {
 		struct i915_buddy_block *block;
-		int order;
+		int max_order = order[i];
+		bool timeout = false;
 		LIST_HEAD(blocks);
+		int order;
 		u64 total;
 
 		err = igt_check_mm(&mm);
@@ -360,6 +368,11 @@ retry:
 			}
 
 			total += i915_buddy_block_size(&mm, block);
+
+			if (__igt_timeout(end_time, NULL)) {
+				timeout = true;
+				break;
+			}
 		} while (total < mm.size);
 
 		if (!err)
@@ -373,7 +386,7 @@ retry:
 				pr_err("post-mm check failed\n");
 		}
 
-		if (err)
+		if (err || timeout)
 			break;
 
 		cond_resched();
@@ -382,6 +395,8 @@ retry:
 	if (err == -ENOMEM)
 		err = 0;
 
+	kfree(order);
+out_fini:
 	i915_buddy_fini(&mm);
 
 	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 34138c7bdd15..0a953bfc0585 100644
--- a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
+++ b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
@@ -43,6 +43,7 @@ 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(ring_submission, intel_ring_submission_live_selftests)
 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/igt_spinner.c b/drivers/gpu/drm/i915/selftests/igt_spinner.c
index e8a58fe49c39..9ad4ab088466 100644
--- a/drivers/gpu/drm/i915/selftests/igt_spinner.c
+++ b/drivers/gpu/drm/i915/selftests/igt_spinner.c
@@ -183,7 +183,7 @@ igt_spinner_create_request(struct igt_spinner *spin,
 
 cancel_rq:
 	if (err) {
-		i915_request_skip(rq, err);
+		i915_request_set_error_once(rq, err);
 		i915_request_add(rq);
 	}
 unpin_hws:
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 3b8986983afc..754d0eb6beaa 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -144,7 +144,6 @@ struct drm_i915_private *mock_gem_device(void)
 		goto put_device;
 	}
 	i915->drm.pdev = pdev;
-	i915->drm.dev_private = i915;
 
 	intel_runtime_pm_init_early(&i915->runtime_pm);