sync with OpenBSD -current

sys/dev/pci/drm/i915/display/i9xx_plane.c

@@ -1059,3 +1059,33 @@ i9xx_get_initial_plane_config(struct intel_crtc *crtc,
 
 	plane_config->fb = intel_fb;
 }
+
+bool i9xx_fixup_initial_plane_config(struct intel_crtc *crtc,
+				     const struct intel_initial_plane_config *plane_config)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct intel_plane *plane = to_intel_plane(crtc->base.primary);
+	const struct intel_plane_state *plane_state =
+		to_intel_plane_state(plane->base.state);
+	enum i9xx_plane_id i9xx_plane = plane->i9xx_plane;
+	u32 base;
+
+	if (!plane_state->uapi.visible)
+		return false;
+
+	base = intel_plane_ggtt_offset(plane_state);
+
+	/*
+	 * We may have moved the surface to a different
+	 * part of ggtt, make the plane aware of that.
+	 */
+	if (plane_config->base == base)
+		return false;
+
+	if (DISPLAY_VER(dev_priv) >= 4)
+		intel_de_write(dev_priv, DSPSURF(i9xx_plane), base);
+	else
+		intel_de_write(dev_priv, DSPADDR(i9xx_plane), base);
+
+	return true;
+}

sys/dev/pci/drm/i915/display/i9xx_plane.h

@@ -26,4 +26,6 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe);
 
 void i9xx_get_initial_plane_config(struct intel_crtc *crtc,
 				   struct intel_initial_plane_config *plane_config);
+bool i9xx_fixup_initial_plane_config(struct intel_crtc *crtc,
+				     const struct intel_initial_plane_config *plane_config);
 #endif

sys/dev/pci/drm/i915/display/intel_display.c

@@ -7764,6 +7764,7 @@ static const struct intel_display_funcs skl_display_funcs = {
 	.crtc_disable = hsw_crtc_disable,
 	.commit_modeset_enables = skl_commit_modeset_enables,
 	.get_initial_plane_config = skl_get_initial_plane_config,
+	.fixup_initial_plane_config = skl_fixup_initial_plane_config,
 };
 
 static const struct intel_display_funcs ddi_display_funcs = {
@@ -7772,6 +7773,7 @@ static const struct intel_display_funcs ddi_display_funcs = {
 	.crtc_disable = hsw_crtc_disable,
 	.commit_modeset_enables = intel_commit_modeset_enables,
 	.get_initial_plane_config = i9xx_get_initial_plane_config,
+	.fixup_initial_plane_config = i9xx_fixup_initial_plane_config,
 };
 
 static const struct intel_display_funcs pch_split_display_funcs = {
@@ -7780,6 +7782,7 @@ static const struct intel_display_funcs pch_split_display_funcs = {
 	.crtc_disable = ilk_crtc_disable,
 	.commit_modeset_enables = intel_commit_modeset_enables,
 	.get_initial_plane_config = i9xx_get_initial_plane_config,
+	.fixup_initial_plane_config = i9xx_fixup_initial_plane_config,
 };
 
 static const struct intel_display_funcs vlv_display_funcs = {
@@ -7788,6 +7791,7 @@ static const struct intel_display_funcs vlv_display_funcs = {
 	.crtc_disable = i9xx_crtc_disable,
 	.commit_modeset_enables = intel_commit_modeset_enables,
 	.get_initial_plane_config = i9xx_get_initial_plane_config,
+	.fixup_initial_plane_config = i9xx_fixup_initial_plane_config,
 };
 
 static const struct intel_display_funcs i9xx_display_funcs = {
@@ -7796,6 +7800,7 @@ static const struct intel_display_funcs i9xx_display_funcs = {
 	.crtc_disable = i9xx_crtc_disable,
 	.commit_modeset_enables = intel_commit_modeset_enables,
 	.get_initial_plane_config = i9xx_get_initial_plane_config,
+	.fixup_initial_plane_config = i9xx_fixup_initial_plane_config,
 };
 
 /**

sys/dev/pci/drm/i915/display/intel_display_core.h

@@ -63,6 +63,8 @@ struct intel_display_funcs {
 				struct intel_crtc_state *);
 	void (*get_initial_plane_config)(struct intel_crtc *,
 					 struct intel_initial_plane_config *);
+	bool (*fixup_initial_plane_config)(struct intel_crtc *crtc,
+					   const struct intel_initial_plane_config *plane_config);
 	void (*crtc_enable)(struct intel_atomic_state *state,
 			    struct intel_crtc *crtc);
 	void (*crtc_disable)(struct intel_atomic_state *state,

sys/dev/pci/drm/i915/display/intel_display_driver.c

@@ -277,7 +277,6 @@ int intel_display_driver_probe_nogem(struct drm_i915_private *i915)
 {
 	struct drm_device *dev = &i915->drm;
 	enum pipe pipe;
-	struct intel_crtc *crtc;
 	int ret;
 
 	if (!HAS_DISPLAY(i915))
@@ -327,11 +326,7 @@ int intel_display_driver_probe_nogem(struct drm_i915_private *i915)
 	intel_acpi_assign_connector_fwnodes(i915);
 	drm_modeset_unlock_all(dev);
 
-	for_each_intel_crtc(dev, crtc) {
-		if (!to_intel_crtc_state(crtc->base.state)->uapi.active)
-			continue;
-		intel_crtc_initial_plane_config(crtc);
-	}
+	intel_initial_plane_config(i915);
 
 	/*
 	 * Make sure hardware watermarks really match the state we read out.

sys/dev/pci/drm/i915/display/intel_display_types.h

@@ -770,6 +770,8 @@ struct intel_plane_state {
 
 struct intel_initial_plane_config {
 	struct intel_framebuffer *fb;
+	struct intel_memory_region *mem;
+	resource_size_t phys_base;
 	struct i915_vma *vma;
 	unsigned int tiling;
 	int size;

sys/dev/pci/drm/i915/display/intel_dpll_mgr.c

@@ -1556,7 +1556,7 @@ static void skl_wrpll_params_populate(struct skl_wrpll_params *params,
 }
 
 static int
-skl_ddi_calculate_wrpll(int clock /* in Hz */,
+skl_ddi_calculate_wrpll(int clock,
 			int ref_clock,
 			struct skl_wrpll_params *wrpll_params)
 {
@@ -1581,7 +1581,7 @@ skl_ddi_calculate_wrpll(int clock /* in Hz */,
 	};
 	unsigned int dco, d, i;
 	unsigned int p0, p1, p2;
-	u64 afe_clock = clock * 5; /* AFE Clock is 5x Pixel clock */
+	u64 afe_clock = (u64)clock * 1000 * 5; /* AFE Clock is 5x Pixel clock, in Hz */
 
 	for (d = 0; d < ARRAY_SIZE(dividers); d++) {
 		for (dco = 0; dco < ARRAY_SIZE(dco_central_freq); dco++) {
@@ -1713,7 +1713,7 @@ static int skl_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state)
 
 	ctrl1 |= DPLL_CTRL1_HDMI_MODE(0);
 
-	ret = skl_ddi_calculate_wrpll(crtc_state->port_clock * 1000,
+	ret = skl_ddi_calculate_wrpll(crtc_state->port_clock,
 				      i915->display.dpll.ref_clks.nssc, &wrpll_params);
 	if (ret)
 		return ret;

sys/dev/pci/drm/i915/display/intel_fbdev.c

@@ -294,8 +294,9 @@ static int intelfb_create(struct drm_fb_helper *helper,
 
 		/* Use fbdev's framebuffer from lmem for discrete */
 		info->fix.smem_start =
-			(unsigned long)(mem->io_start +
-					i915_gem_object_get_dma_address(obj, 0));
+			(unsigned long)(mem->io.start +
+					i915_gem_object_get_dma_address(obj, 0) -
+					mem->region.start);
 		info->fix.smem_len = obj->base.size;
 	} else {
 		/* Our framebuffer is the entirety of fbdev's system memory */

sys/dev/pci/drm/i915/display/intel_hdcp_regs.h

@@ -249,7 +249,7 @@
 #define HDCP2_STREAM_STATUS(dev_priv, trans, port) \
 					(GRAPHICS_VER(dev_priv) >= 12 ? \
 					 TRANS_HDCP2_STREAM_STATUS(trans) : \
-					 PIPE_HDCP2_STREAM_STATUS(pipe))
+					 PIPE_HDCP2_STREAM_STATUS(port))
 
 #define _PORTA_HDCP2_AUTH_STREAM		0x66F00
 #define _PORTB_HDCP2_AUTH_STREAM		0x66F04

sys/dev/pci/drm/i915/display/intel_plane_initial.c

@@ -3,29 +3,32 @@
  * Copyright © 2021 Intel Corporation
  */
 
+#include "gem/i915_gem_lmem.h"
 #include "gem/i915_gem_region.h"
 #include "i915_drv.h"
 #include "intel_atomic_plane.h"
+#include "intel_crtc.h"
 #include "intel_display.h"
 #include "intel_display_types.h"
 #include "intel_fb.h"
 #include "intel_plane_initial.h"
 
 static bool
-intel_reuse_initial_plane_obj(struct drm_i915_private *i915,
-			      const struct intel_initial_plane_config *plane_config,
+intel_reuse_initial_plane_obj(struct intel_crtc *this,
+			      const struct intel_initial_plane_config plane_configs[],
 			      struct drm_framebuffer **fb,
 			      struct i915_vma **vma)
 {
+	struct drm_i915_private *i915 = to_i915(this->base.dev);
 	struct intel_crtc *crtc;
 
 	for_each_intel_crtc(&i915->drm, crtc) {
-		struct intel_crtc_state *crtc_state =
-			to_intel_crtc_state(crtc->base.state);
 		struct intel_plane *plane =
 			to_intel_plane(crtc->base.primary);
-		struct intel_plane_state *plane_state =
+		const struct intel_plane_state *plane_state =
 			to_intel_plane_state(plane->base.state);
+		const struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
 
 		if (!crtc_state->uapi.active)
 			continue;
@@ -33,7 +36,7 @@ intel_reuse_initial_plane_obj(struct drm_i915_private *i915,
 		if (!plane_state->ggtt_vma)
 			continue;
 
-		if (intel_plane_ggtt_offset(plane_state) == plane_config->base) {
+		if (plane_configs[this->pipe].base == plane_configs[crtc->pipe].base) {
 			*fb = plane_state->hw.fb;
 			*vma = plane_state->ggtt_vma;
 			return true;
@@ -43,12 +46,100 @@ intel_reuse_initial_plane_obj(struct drm_i915_private *i915,
 	return false;
 }
 
+static bool
+initial_plane_phys_lmem(struct drm_i915_private *i915,
+			struct intel_initial_plane_config *plane_config)
+{
+	gen8_pte_t __iomem *gte = to_gt(i915)->ggtt->gsm;
+	struct intel_memory_region *mem;
+	dma_addr_t dma_addr;
+	gen8_pte_t pte;
+	u32 base;
+
+	base = round_down(plane_config->base, I915_GTT_MIN_ALIGNMENT);
+
+	gte += base / I915_GTT_PAGE_SIZE;
+
+	pte = ioread64(gte);
+	if (!(pte & GEN12_GGTT_PTE_LM)) {
+		drm_err(&i915->drm,
+			"Initial plane programming missing PTE_LM bit\n");
+		return false;
+	}
+
+	dma_addr = pte & GEN12_GGTT_PTE_ADDR_MASK;
+
+	if (IS_DGFX(i915))
+		mem = i915->mm.regions[INTEL_REGION_LMEM_0];
+	else
+		mem = i915->mm.stolen_region;
+	if (!mem) {
+		drm_dbg_kms(&i915->drm,
+			    "Initial plane memory region not initialized\n");
+		return false;
+	}
+
+	/*
+	 * On lmem we don't currently expect this to
+	 * ever be placed in the stolen portion.
+	 */
+	if (dma_addr < mem->region.start || dma_addr > mem->region.end) {
+		drm_err(&i915->drm,
+			"Initial plane programming using invalid range, dma_addr=%pa (%s [%pa-%pa])\n",
+			&dma_addr, mem->region.name, &mem->region.start, &mem->region.end);
+		return false;
+	}
+
+	drm_dbg(&i915->drm,
+		"Using dma_addr=%pa, based on initial plane programming\n",
+		&dma_addr);
+
+	plane_config->phys_base = dma_addr - mem->region.start;
+	plane_config->mem = mem;
+
+	return true;
+}
+
+static bool
+initial_plane_phys_smem(struct drm_i915_private *i915,
+			struct intel_initial_plane_config *plane_config)
+{
+	struct intel_memory_region *mem;
+	u32 base;
+
+	base = round_down(plane_config->base, I915_GTT_MIN_ALIGNMENT);
+
+	mem = i915->mm.stolen_region;
+	if (!mem) {
+		drm_dbg_kms(&i915->drm,
+			    "Initial plane memory region not initialized\n");
+		return false;
+	}
+
+	/* FIXME get and validate the dma_addr from the PTE */
+	plane_config->phys_base = base;
+	plane_config->mem = mem;
+
+	return true;
+}
+
+static bool
+initial_plane_phys(struct drm_i915_private *i915,
+		   struct intel_initial_plane_config *plane_config)
+{
+	if (IS_DGFX(i915) || HAS_LMEMBAR_SMEM_STOLEN(i915))
+		return initial_plane_phys_lmem(i915, plane_config);
+	else
+		return initial_plane_phys_smem(i915, plane_config);
+}
+
 static struct i915_vma *
 initial_plane_vma(struct drm_i915_private *i915,
 		  struct intel_initial_plane_config *plane_config)
 {
 	struct intel_memory_region *mem;
 	struct drm_i915_gem_object *obj;
+	struct drm_mm_node orig_mm = {};
 	struct i915_vma *vma;
 	resource_size_t phys_base;
 	u32 base, size;
@@ -57,45 +148,13 @@ initial_plane_vma(struct drm_i915_private *i915,
 	if (plane_config->size == 0)
 		return NULL;
 
-	base = round_down(plane_config->base, I915_GTT_MIN_ALIGNMENT);
-	if (IS_DGFX(i915)) {
-		gen8_pte_t __iomem *gte = to_gt(i915)->ggtt->gsm;
-		gen8_pte_t pte;
-
-		gte += base / I915_GTT_PAGE_SIZE;
-
-		pte = ioread64(gte);
-		if (!(pte & GEN12_GGTT_PTE_LM)) {
-			drm_err(&i915->drm,
-				"Initial plane programming missing PTE_LM bit\n");
-			return NULL;
-		}
-
-		phys_base = pte & I915_GTT_PAGE_MASK;
-		mem = i915->mm.regions[INTEL_REGION_LMEM_0];
-
-		/*
-		 * We don't currently expect this to ever be placed in the
-		 * stolen portion.
-		 */
-		if (phys_base >= resource_size(&mem->region)) {
-			drm_err(&i915->drm,
-				"Initial plane programming using invalid range, phys_base=%pa\n",
-				&phys_base);
-			return NULL;
-		}
-
-		drm_dbg(&i915->drm,
-			"Using phys_base=%pa, based on initial plane programming\n",
-			&phys_base);
-	} else {
-		phys_base = base;
-		mem = i915->mm.stolen_region;
-	}
-
-	if (!mem)
+	if (!initial_plane_phys(i915, plane_config))
 		return NULL;
 
+	phys_base = plane_config->phys_base;
+	mem = plane_config->mem;
+
+	base = round_down(plane_config->base, I915_GTT_MIN_ALIGNMENT);
 	size = round_up(plane_config->base + plane_config->size,
 			mem->min_page_size);
 	size -= base;
@@ -107,14 +166,19 @@ initial_plane_vma(struct drm_i915_private *i915,
 	 */
 	if (IS_ENABLED(CONFIG_FRAMEBUFFER_CONSOLE) &&
 	    mem == i915->mm.stolen_region &&
-	    size * 2 > i915->dsm.usable_size)
+	    size * 2 > i915->dsm.usable_size) {
+		drm_dbg_kms(&i915->drm, "Initial FB size exceeds half of stolen, discarding\n");
 		return NULL;
+	}
 
 	obj = i915_gem_object_create_region_at(mem, phys_base, size,
 					       I915_BO_ALLOC_USER |
 					       I915_BO_PREALLOC);
-	if (IS_ERR(obj))
+	if (IS_ERR(obj)) {
+		drm_dbg_kms(&i915->drm, "Failed to preallocate initial FB in %s\n",
+			    mem->region.name);
 		return NULL;
+	}
 
 	/*
 	 * Mark it WT ahead of time to avoid changing the
@@ -138,23 +202,66 @@ initial_plane_vma(struct drm_i915_private *i915,
 		goto err_obj;
 	}
 
+	/*
+	 * MTL GOP likes to place the framebuffer high up in ggtt,
+	 * which can cause problems for ggtt_reserve_guc_top().
+	 * Try to pin it to a low ggtt address instead to avoid that.
+	 */
+	base = 0;
+
+	if (base != plane_config->base) {
+		struct i915_ggtt *ggtt = to_gt(i915)->ggtt;
+		int ret;
+
+		/*
+		 * Make sure the original and new locations
+		 * can't overlap. That would corrupt the original
+		 * PTEs which are still being used for scanout.
+		 */
+		ret = i915_gem_gtt_reserve(&ggtt->vm, NULL, &orig_mm,
+					   size, plane_config->base,
+					   I915_COLOR_UNEVICTABLE, PIN_NOEVICT);
+		if (ret)
+			goto err_obj;
+	}
+
 	vma = i915_vma_instance(obj, &to_gt(i915)->ggtt->vm, NULL);
 	if (IS_ERR(vma))
 		goto err_obj;
 
+retry:
 	pinctl = PIN_GLOBAL | PIN_OFFSET_FIXED | base;
-	if (HAS_GMCH(i915))
+	if (!i915_gem_object_is_lmem(obj))
 		pinctl |= PIN_MAPPABLE;
-	if (i915_vma_pin(vma, 0, 0, pinctl))
+	if (i915_vma_pin(vma, 0, 0, pinctl)) {
+		if (drm_mm_node_allocated(&orig_mm)) {
+			drm_mm_remove_node(&orig_mm);
+			/*
+			 * Try again, but this time pin
+			 * it to its original location.
+			 */
+			base = plane_config->base;
+			goto retry;
+		}
 		goto err_obj;
+	}
 
 	if (i915_gem_object_is_tiled(obj) &&
 	    !i915_vma_is_map_and_fenceable(vma))
 		goto err_obj;
 
+	if (drm_mm_node_allocated(&orig_mm))
+		drm_mm_remove_node(&orig_mm);
+
+	drm_dbg_kms(&i915->drm,
+		    "Initial plane fb bound to 0x%x in the ggtt (original 0x%x)\n",
+		    i915_ggtt_offset(vma), plane_config->base);
+
 	return vma;
 
 err_obj:
+	if (drm_mm_node_allocated(&orig_mm))
+		drm_mm_remove_node(&orig_mm);
 	i915_gem_object_put(obj);
 	return NULL;
 }
@@ -209,10 +316,11 @@ err_vma:
 
 static void
 intel_find_initial_plane_obj(struct intel_crtc *crtc,
-			     struct intel_initial_plane_config *plane_config)
+			     struct intel_initial_plane_config plane_configs[])
 {
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	struct intel_initial_plane_config *plane_config =
+		&plane_configs[crtc->pipe];
 	struct intel_plane *plane =
 		to_intel_plane(crtc->base.primary);
 	struct intel_plane_state *plane_state =
@@ -238,7 +346,7 @@ intel_find_initial_plane_obj(struct intel_crtc *crtc,
 	 * Failed to alloc the obj, check to see if we should share
 	 * an fb with another CRTC instead
 	 */
-	if (intel_reuse_initial_plane_obj(dev_priv, plane_config, &fb, &vma))
+	if (intel_reuse_initial_plane_obj(crtc, plane_configs, &fb, &vma))
 		goto valid_fb;
 
 	/*
@@ -301,25 +409,36 @@ static void plane_config_fini(struct intel_initial_plane_config *plane_config)
 		i915_vma_put(plane_config->vma);
 }
 
-void intel_crtc_initial_plane_config(struct intel_crtc *crtc)
+void intel_initial_plane_config(struct drm_i915_private *i915)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	struct intel_initial_plane_config plane_config = {};
+	struct intel_initial_plane_config plane_configs[I915_MAX_PIPES] = {};
+	struct intel_crtc *crtc;
 
-	/*
-	 * Note that reserving the BIOS fb up front prevents us
-	 * from stuffing other stolen allocations like the ring
-	 * on top.  This prevents some ugliness at boot time, and
-	 * can even allow for smooth boot transitions if the BIOS
-	 * fb is large enough for the active pipe configuration.
-	 */
-	dev_priv->display.funcs.display->get_initial_plane_config(crtc, &plane_config);
+	for_each_intel_crtc(&i915->drm, crtc) {
+		struct intel_initial_plane_config *plane_config =
+			&plane_configs[crtc->pipe];
 
-	/*
-	 * If the fb is shared between multiple heads, we'll
-	 * just get the first one.
-	 */
-	intel_find_initial_plane_obj(crtc, &plane_config);
+		if (!to_intel_crtc_state(crtc->base.state)->uapi.active)
+			continue;
 
-	plane_config_fini(&plane_config);
+		/*
+		 * Note that reserving the BIOS fb up front prevents us
+		 * from stuffing other stolen allocations like the ring
+		 * on top.  This prevents some ugliness at boot time, and
+		 * can even allow for smooth boot transitions if the BIOS
+		 * fb is large enough for the active pipe configuration.
+		 */
+		i915->display.funcs.display->get_initial_plane_config(crtc, plane_config);
+
+		/*
+		 * If the fb is shared between multiple heads, we'll
+		 * just get the first one.
+		 */
+		intel_find_initial_plane_obj(crtc, plane_configs);
+
+		if (i915->display.funcs.display->fixup_initial_plane_config(crtc, plane_config))
+			intel_crtc_wait_for_next_vblank(crtc);
+
+		plane_config_fini(plane_config);
+	}
 }

sys/dev/pci/drm/i915/display/intel_plane_initial.h

@@ -6,8 +6,8 @@
 #ifndef __INTEL_PLANE_INITIAL_H__
 #define __INTEL_PLANE_INITIAL_H__
 
-struct intel_crtc;
+struct drm_i915_private;
 
-void intel_crtc_initial_plane_config(struct intel_crtc *crtc);
+void intel_initial_plane_config(struct drm_i915_private *i915);
 
 #endif

sys/dev/pci/drm/i915/display/skl_universal_plane.c

@@ -2557,3 +2557,31 @@ skl_get_initial_plane_config(struct intel_crtc *crtc,
 error:
 	kfree(intel_fb);
 }
+
+bool skl_fixup_initial_plane_config(struct intel_crtc *crtc,
+				    const struct intel_initial_plane_config *plane_config)
+{
+	struct drm_i915_private *i915 = to_i915(crtc->base.dev);
+	struct intel_plane *plane = to_intel_plane(crtc->base.primary);
+	const struct intel_plane_state *plane_state =
+		to_intel_plane_state(plane->base.state);
+	enum plane_id plane_id = plane->id;
+	enum pipe pipe = crtc->pipe;
+	u32 base;
+
+	if (!plane_state->uapi.visible)
+		return false;
+
+	base = intel_plane_ggtt_offset(plane_state);
+
+	/*
+	 * We may have moved the surface to a different
+	 * part of ggtt, make the plane aware of that.
+	 */
+	if (plane_config->base == base)
+		return false;
+
+	intel_de_write(i915, PLANE_SURF(pipe, plane_id), base);
+
+	return true;
+}

sys/dev/pci/drm/i915/display/skl_universal_plane.h

@@ -22,6 +22,8 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 
 void skl_get_initial_plane_config(struct intel_crtc *crtc,
 				  struct intel_initial_plane_config *plane_config);
+bool skl_fixup_initial_plane_config(struct intel_crtc *crtc,
+				    const struct intel_initial_plane_config *plane_config);
 
 int skl_format_to_fourcc(int format, bool rgb_order, bool alpha);
 

sys/dev/pci/drm/i915/gem/i915_gem_region.c

@@ -129,7 +129,7 @@ i915_gem_object_create_region_at(struct intel_memory_region *mem,
 		return ERR_PTR(-EINVAL);
 
 	if (!(flags & I915_BO_ALLOC_GPU_ONLY) &&
-	    offset + size > mem->io_size &&
+	    offset + size > resource_size(&mem->io) &&
 	    !i915_ggtt_has_aperture(to_gt(mem->i915)->ggtt))
 		return ERR_PTR(-ENOSPC);
 

sys/dev/pci/drm/i915/gem/i915_gem_stolen.c

@@ -545,7 +545,9 @@ static int i915_gem_init_stolen(struct intel_memory_region *mem)
 
 	/* Exclude the reserved region from driver use */
 	mem->region.end = i915->dsm.reserved.start - 1;
-	mem->io_size = min(mem->io_size, resource_size(&mem->region));
+	mem->io = DEFINE_RES_MEM(mem->io.start,
+				 min(resource_size(&mem->io),
+				     resource_size(&mem->region)));
 
 	i915->dsm.usable_size = resource_size(&mem->region);
 
@@ -756,7 +758,7 @@ static int _i915_gem_object_stolen_init(struct intel_memory_region *mem,
 	 * With discrete devices, where we lack a mappable aperture there is no
 	 * possible way to ever access this memory on the CPU side.
 	 */
-	if (mem->type == INTEL_MEMORY_STOLEN_LOCAL && !mem->io_size &&
+	if (mem->type == INTEL_MEMORY_STOLEN_LOCAL && !resource_size(&mem->io) &&
 	    !(flags & I915_BO_ALLOC_GPU_ONLY))
 		return -ENOSPC;
 
@@ -830,7 +832,6 @@ static const struct intel_memory_region_ops i915_region_stolen_smem_ops = {
 
 static int init_stolen_lmem(struct intel_memory_region *mem)
 {
-	struct drm_i915_private *i915 = mem->i915;
 	int err;
 
 	if (GEM_WARN_ON(resource_size(&mem->region) == 0))
@@ -843,38 +844,34 @@ static int init_stolen_lmem(struct intel_memory_region *mem)
 	}
 
 #ifdef __linux__
-	if (mem->io_size &&
-	    !io_mapping_init_wc(&mem->iomap, mem->io_start, mem->io_size))
+	if (resource_size(&mem->io) &&
+	    !io_mapping_init_wc(&mem->iomap, mem->io.start, resource_size(&mem->io)))
 		goto err_cleanup;
 #else
-	if (mem->io_size) {
+	if (resource_size(&mem->io)) {
 		paddr_t start, end;
 		struct vm_page *pgs;
 		int i;
 		bus_space_handle_t bsh;
 
-		start = atop(mem->io_start);
-		end = start + atop(mem->io_size);
+		start = atop(mem->io.start);
+		end = start + atop(resource_size(&mem->io));
 		uvm_page_physload(start, end, start, end, PHYSLOAD_DEVICE);
 
-		pgs = PHYS_TO_VM_PAGE(mem->io_start);
-		for (i = 0; i < atop(mem->io_size); i++)
+		pgs = PHYS_TO_VM_PAGE(mem->io.start);
+		for (i = 0; i < atop(resource_size(&mem->io)); i++)
 			atomic_setbits_int(&(pgs[i].pg_flags), PG_PMAP_WC);
 
-		if (bus_space_map(i915->bst, mem->io_start, mem->io_size,
+		if (bus_space_map(mem->i915->bst, mem->io.start, resource_size(&mem->io),
 		    BUS_SPACE_MAP_LINEAR | BUS_SPACE_MAP_PREFETCHABLE, &bsh))
 			panic("can't map stolen lmem");
 
-		mem->iomap.base = mem->io_start;
-		mem->iomap.size = mem->io_size;
-		mem->iomap.iomem = bus_space_vaddr(i915->bst, bsh);
+		mem->iomap.base = mem->io.start;
+		mem->iomap.size = resource_size(&mem->io);
+		mem->iomap.iomem = bus_space_vaddr(mem->i915->bst, bsh);
 	}
 #endif
 
-	drm_dbg(&i915->drm, "Stolen Local memory IO start: %pa\n",
-		&mem->io_start);
-	drm_dbg(&i915->drm, "Stolen Local DSM base: %pa\n", &mem->region.start);
-
 	return 0;
 #ifdef __linux__
 err_cleanup:
@@ -887,7 +884,7 @@ static int release_stolen_lmem(struct intel_memory_region *mem)
 {
 	STUB();
 #ifdef notyet
-	if (mem->io_size)
+	if (resource_size(&mem->io))
 		io_mapping_fini(&mem->iomap);
 #endif
 	i915_gem_cleanup_stolen(mem->i915);
@@ -987,14 +984,18 @@ i915_gem_stolen_lmem_setup(struct drm_i915_private *i915, u16 type,
 		GEM_BUG_ON((dsm_base + dsm_size) > lmem_size);
 	} else {
 		/* Use DSM base address instead for stolen memory */
-		dsm_base = intel_uncore_read64(uncore, GEN12_DSMBASE) & GEN12_BDSM_MASK;
+		dsm_base = intel_uncore_read64(uncore, GEN6_DSMBASE) & GEN11_BDSM_MASK;
 		if (WARN_ON(lmem_size < dsm_base))
 			return ERR_PTR(-ENODEV);
 		dsm_size = ALIGN_DOWN(lmem_size - dsm_base, SZ_1M);
 	}
 
+	if (i915_direct_stolen_access(i915)) {
+		drm_dbg(&i915->drm, "Using direct DSM access\n");
+		io_start = intel_uncore_read64(uncore, GEN6_DSMBASE) & GEN11_BDSM_MASK;
+		io_size = dsm_size;
 #ifdef __linux__
-	if (pci_resource_len(pdev, GEN12_LMEM_BAR) < lmem_size) {
+	} else if (pci_resource_len(pdev, GEN12_LMEM_BAR) < lmem_size) {
 		io_start = 0;
 		io_size = 0;
 	} else {
@@ -1002,7 +1003,7 @@ i915_gem_stolen_lmem_setup(struct drm_i915_private *i915, u16 type,
 		io_size = dsm_size;
 	}
 #else
-	if (lmem_len < lmem_size) {
+	} else if (lmem_len < lmem_size) {
 		io_start = 0;
 		io_size = 0;
 	} else {

sys/dev/pci/drm/i915/gem/i915_gem_ttm.c

@@ -144,13 +144,13 @@ i915_ttm_place_from_region(const struct intel_memory_region *mr,
 		place->fpfn = offset >> PAGE_SHIFT;
 		WARN_ON(overflows_type(place->fpfn + (size >> PAGE_SHIFT), place->lpfn));
 		place->lpfn = place->fpfn + (size >> PAGE_SHIFT);
-	} else if (mr->io_size && mr->io_size < mr->total) {
+	} else if (resource_size(&mr->io) && resource_size(&mr->io) < mr->total) {
 		if (flags & I915_BO_ALLOC_GPU_ONLY) {
 			place->flags |= TTM_PL_FLAG_TOPDOWN;
 		} else {
 			place->fpfn = 0;
-			WARN_ON(overflows_type(mr->io_size >> PAGE_SHIFT, place->lpfn));
-			place->lpfn = mr->io_size >> PAGE_SHIFT;
+			WARN_ON(overflows_type(resource_size(&mr->io) >> PAGE_SHIFT, place->lpfn));
+			place->lpfn = resource_size(&mr->io) >> PAGE_SHIFT;
 		}
 	}
 }
@@ -1121,7 +1121,7 @@ static vm_fault_t vm_fault_ttm(struct vm_fault *vmf)
 			struct intel_memory_region *mr = obj->mm.placements[i];
 			unsigned int flags;
 
-			if (!mr->io_size && mr->type != INTEL_MEMORY_SYSTEM)
+			if (!resource_size(&mr->io) && mr->type != INTEL_MEMORY_SYSTEM)
 				continue;
 
 			flags = obj->flags;
@@ -1295,7 +1295,7 @@ vm_fault_ttm(struct uvm_faultinfo *ufi, vaddr_t vaddr, vm_page_t *pps,
 			struct intel_memory_region *mr = obj->mm.placements[i];
 			unsigned int flags;
 
-			if (!mr->io_size && mr->type != INTEL_MEMORY_SYSTEM)
+			if (!resource_size(&mr->io) && mr->type != INTEL_MEMORY_SYSTEM)
 				continue;
 
 			flags = obj->flags;

sys/dev/pci/drm/i915/gem/selftests/i915_gem_mman.c

@@ -1052,7 +1052,7 @@ static int igt_fill_mappable(struct intel_memory_region *mr,
 	int err;
 
 	total = 0;
-	size = mr->io_size;
+	size = resource_size(&mr->io);
 	do {
 		struct drm_i915_gem_object *obj;
 
@@ -1313,28 +1313,28 @@ static int igt_mmap_migrate(void *arg)
 		struct intel_memory_region *mixed[] = { mr, system };
 		struct intel_memory_region *single[] = { mr };
 		struct ttm_resource_manager *man = mr->region_private;
-		resource_size_t saved_io_size;
+		struct resource saved_io;
 		int err;
 
 		if (mr->private)
 			continue;
 
-		if (!mr->io_size)
+		if (!resource_size(&mr->io))
 			continue;
 
 		/*
 		 * For testing purposes let's force small BAR, if not already
 		 * present.
 		 */
-		saved_io_size = mr->io_size;
-		if (mr->io_size == mr->total) {
-			resource_size_t io_size = mr->io_size;
+		saved_io = mr->io;
+		if (resource_size(&mr->io) == mr->total) {
+			resource_size_t io_size = resource_size(&mr->io);
 
 			io_size = rounddown_pow_of_two(io_size >> 1);
 			if (io_size < PAGE_SIZE)
 				continue;
 
-			mr->io_size = io_size;
+			mr->io = DEFINE_RES_MEM(mr->io.start, io_size);
 			i915_ttm_buddy_man_force_visible_size(man,
 							      io_size >> PAGE_SHIFT);
 		}
@@ -1394,9 +1394,9 @@ static int igt_mmap_migrate(void *arg)
 					 IGT_MMAP_MIGRATE_FAIL_GPU |
 					 IGT_MMAP_MIGRATE_UNFAULTABLE);
 out_io_size:
-		mr->io_size = saved_io_size;
+		mr->io = saved_io;
 		i915_ttm_buddy_man_force_visible_size(man,
-						      mr->io_size >> PAGE_SHIFT);
+						      resource_size(&mr->io) >> PAGE_SHIFT);
 		if (err)
 			return err;
 	}

sys/dev/pci/drm/i915/gt/gen8_engine_cs.c

@@ -278,7 +278,8 @@ int gen12_emit_flush_rcs(struct i915_request *rq, u32 mode)
 		 * deals with Protected Memory which is not needed for
 		 * AUX CCS invalidation and lead to unwanted side effects.
 		 */
-		if (mode & EMIT_FLUSH)
+		if ((mode & EMIT_FLUSH) &&
+		    GRAPHICS_VER_FULL(rq->i915) < IP_VER(12, 70))
 			bit_group_1 |= PIPE_CONTROL_FLUSH_L3;
 
 		bit_group_1 |= PIPE_CONTROL_TILE_CACHE_FLUSH;
@@ -812,12 +813,14 @@ u32 *gen12_emit_fini_breadcrumb_rcs(struct i915_request *rq, u32 *cs)
 	u32 flags = (PIPE_CONTROL_CS_STALL |
 		     PIPE_CONTROL_TLB_INVALIDATE |
 		     PIPE_CONTROL_TILE_CACHE_FLUSH |
-		     PIPE_CONTROL_FLUSH_L3 |
 		     PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH |
 		     PIPE_CONTROL_DEPTH_CACHE_FLUSH |
 		     PIPE_CONTROL_DC_FLUSH_ENABLE |
 		     PIPE_CONTROL_FLUSH_ENABLE);
 
+	if (GRAPHICS_VER_FULL(rq->i915) < IP_VER(12, 70))
+		flags |= PIPE_CONTROL_FLUSH_L3;
+
 	/* Wa_14016712196 */
 	if (IS_GFX_GT_IP_RANGE(gt, IP_VER(12, 70), IP_VER(12, 74)) || IS_DG2(i915))
 		/* dummy PIPE_CONTROL + depth flush */

sys/dev/pci/drm/i915/gt/intel_engine.h

@@ -170,6 +170,8 @@ intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value)
 #define I915_GEM_HWS_SEQNO		0x40
 #define I915_GEM_HWS_SEQNO_ADDR		(I915_GEM_HWS_SEQNO * sizeof(u32))
 #define I915_GEM_HWS_MIGRATE		(0x42 * sizeof(u32))
+#define I915_GEM_HWS_GGTT_BIND		0x46
+#define I915_GEM_HWS_GGTT_BIND_ADDR	(I915_GEM_HWS_GGTT_BIND * sizeof(u32))
 #define I915_GEM_HWS_PXP		0x60
 #define I915_GEM_HWS_PXP_ADDR		(I915_GEM_HWS_PXP * sizeof(u32))
 #define I915_GEM_HWS_GSC		0x62

sys/dev/pci/drm/i915/gt/intel_engine_cs.c

@@ -1445,6 +1445,20 @@ void intel_engine_destroy_pinned_context(struct intel_context *ce)
 	intel_context_put(ce);
 }
 
+static struct intel_context *
+create_ggtt_bind_context(struct intel_engine_cs *engine)
+{
+	static struct lock_class_key kernel;
+
+	/*
+	 * MI_UPDATE_GTT can insert up to 511 PTE entries and there could be multiple
+	 * bind requets at a time so get a bigger ring.
+	 */
+	return intel_engine_create_pinned_context(engine, engine->gt->vm, SZ_512K,
+						  I915_GEM_HWS_GGTT_BIND_ADDR,
+						  &kernel, "ggtt_bind_context");
+}
+
 static struct intel_context *
 create_kernel_context(struct intel_engine_cs *engine)
 {
@@ -1468,7 +1482,7 @@ create_kernel_context(struct intel_engine_cs *engine)
  */
 static int engine_init_common(struct intel_engine_cs *engine)
 {
-	struct intel_context *ce;
+	struct intel_context *ce, *bce = NULL;
 	int ret;
 
 	engine->set_default_submission(engine);
@@ -1484,17 +1498,34 @@ static int engine_init_common(struct intel_engine_cs *engine)
 	ce = create_kernel_context(engine);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
+	/*
+	 * Create a separate pinned context for GGTT update with blitter engine
+	 * if a platform require such service. MI_UPDATE_GTT works on other
+	 * engines as well but BCS should be less busy engine so pick that for
+	 * GGTT updates.
+	 */
+	if (i915_ggtt_require_binder(engine->i915) && engine->id == BCS0) {
+		bce = create_ggtt_bind_context(engine);
+		if (IS_ERR(bce)) {
+			ret = PTR_ERR(bce);
+			goto err_ce_context;
+		}
+	}
 
 	ret = measure_breadcrumb_dw(ce);
 	if (ret < 0)
-		goto err_context;
+		goto err_bce_context;
 
 	engine->emit_fini_breadcrumb_dw = ret;
 	engine->kernel_context = ce;
+	engine->bind_context = bce;
 
 	return 0;
 
-err_context:
+err_bce_context:
+	if (bce)
+		intel_engine_destroy_pinned_context(bce);
+err_ce_context:
 	intel_engine_destroy_pinned_context(ce);
 	return ret;
 }
@@ -1564,6 +1595,10 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 	if (engine->kernel_context)
 		intel_engine_destroy_pinned_context(engine->kernel_context);
 
+	if (engine->bind_context)
+		intel_engine_destroy_pinned_context(engine->bind_context);
+
+
 	GEM_BUG_ON(!llist_empty(&engine->barrier_tasks));
 	cleanup_status_page(engine);
 

sys/dev/pci/drm/i915/gt/intel_engine_types.h

@@ -416,6 +416,9 @@ struct intel_engine_cs {
 	struct llist_head barrier_tasks;
 
 	struct intel_context *kernel_context; /* pinned */
+	struct intel_context *bind_context; /* pinned, only for BCS0 */
+	/* mark the bind context's availability status */
+	bool bind_context_ready;
 
 	/**
 	 * pinned_contexts_list: List of pinned contexts. This list is only

sys/dev/pci/drm/i915/gt/intel_ggtt.c

@@ -15,18 +15,24 @@
 #include "display/intel_display.h"
 #include "gem/i915_gem_lmem.h"
 
+#include "intel_context.h"
 #include "intel_ggtt_gmch.h"
+#include "intel_gpu_commands.h"
 #include "intel_gt.h"
 #include "intel_gt_regs.h"
 #include "intel_pci_config.h"
+#include "intel_ring.h"
 #include "i915_drv.h"
 #include "i915_pci.h"
+#include "i915_reg.h"
+#include "i915_request.h"
 #include "i915_scatterlist.h"
 #include "i915_utils.h"
 #include "i915_vgpu.h"
 
 #include "intel_gtt.h"
 #include "gen8_ppgtt.h"
+#include "intel_engine_pm.h"
 
 static void i915_ggtt_color_adjust(const struct drm_mm_node *node,
 				   unsigned long color,
@@ -301,6 +307,145 @@ u64 gen8_ggtt_pte_encode(dma_addr_t addr,
 	return pte;
 }
 
+static bool should_update_ggtt_with_bind(struct i915_ggtt *ggtt)
+{
+	struct intel_gt *gt = ggtt->vm.gt;
+
+	return intel_gt_is_bind_context_ready(gt);
+}
+
+static struct intel_context *gen8_ggtt_bind_get_ce(struct i915_ggtt *ggtt)
+{
+	struct intel_context *ce;
+	struct intel_gt *gt = ggtt->vm.gt;
+
+	if (intel_gt_is_wedged(gt))
+		return NULL;
+
+	ce = gt->engine[BCS0]->bind_context;
+	GEM_BUG_ON(!ce);
+
+	/*
+	 * If the GT is not awake already at this stage then fallback
+	 * to pci based GGTT update otherwise __intel_wakeref_get_first()
+	 * would conflict with fs_reclaim trying to allocate memory while
+	 * doing rpm_resume().
+	 */
+	if (!intel_gt_pm_get_if_awake(gt))
+		return NULL;
+
+	intel_engine_pm_get(ce->engine);
+
+	return ce;
+}
+
+static void gen8_ggtt_bind_put_ce(struct intel_context *ce)
+{
+	intel_engine_pm_put(ce->engine);
+	intel_gt_pm_put(ce->engine->gt);
+}
+
+static bool gen8_ggtt_bind_ptes(struct i915_ggtt *ggtt, u32 offset,
+				struct sg_table *pages, u32 num_entries,
+				const gen8_pte_t pte)
+{
+	struct i915_sched_attr attr = {};
+	struct intel_gt *gt = ggtt->vm.gt;
+	const gen8_pte_t scratch_pte = ggtt->vm.scratch[0]->encode;
+	struct sgt_iter iter;
+	struct i915_request *rq;
+	struct intel_context *ce;
+	u32 *cs;
+
+	if (!num_entries)
+		return true;
+
+	ce = gen8_ggtt_bind_get_ce(ggtt);
+	if (!ce)
+		return false;
+
+	if (pages)
+		iter = __sgt_iter(pages->sgl, true);
+
+	while (num_entries) {
+		int count = 0;
+		dma_addr_t addr;
+		/*
+		 * MI_UPDATE_GTT can update 512 entries in a single command but
+		 * that end up with engine reset, 511 works.
+		 */
+		u32 n_ptes = min_t(u32, 511, num_entries);
+
+		if (mutex_lock_interruptible(&ce->timeline->mutex))
+			goto put_ce;
+
+		intel_context_enter(ce);
+		rq = __i915_request_create(ce, GFP_NOWAIT | GFP_ATOMIC);
+		intel_context_exit(ce);
+		if (IS_ERR(rq)) {
+			GT_TRACE(gt, "Failed to get bind request\n");
+			mutex_unlock(&ce->timeline->mutex);
+			goto put_ce;
+		}
+
+		cs = intel_ring_begin(rq, 2 * n_ptes + 2);
+		if (IS_ERR(cs)) {
+			GT_TRACE(gt, "Failed to ring space for GGTT bind\n");
+			i915_request_set_error_once(rq, PTR_ERR(cs));
+			/* once a request is created, it must be queued */
+			goto queue_err_rq;
+		}
+
+		*cs++ = MI_UPDATE_GTT | (2 * n_ptes);
+		*cs++ = offset << 12;
+
+		if (pages) {
+			for_each_sgt_daddr_next(addr, iter) {
+				if (count == n_ptes)
+					break;
+				*cs++ = lower_32_bits(pte | addr);
+				*cs++ = upper_32_bits(pte | addr);
+				count++;
+			}
+			/* fill remaining with scratch pte, if any */
+			if (count < n_ptes) {
+				memset64((u64 *)cs, scratch_pte,
+					 n_ptes - count);
+				cs += (n_ptes - count) * 2;
+			}
+		} else {
+			memset64((u64 *)cs, pte, n_ptes);
+			cs += n_ptes * 2;
+		}
+
+		intel_ring_advance(rq, cs);
+queue_err_rq:
+		i915_request_get(rq);
+		__i915_request_commit(rq);
+		__i915_request_queue(rq, &attr);
+
+		mutex_unlock(&ce->timeline->mutex);
+		/* This will break if the request is complete or after engine reset */
+		i915_request_wait(rq, 0, MAX_SCHEDULE_TIMEOUT);
+		if (rq->fence.error)
+			goto err_rq;
+
+		i915_request_put(rq);
+
+		num_entries -= n_ptes;
+		offset += n_ptes;
+	}
+
+	gen8_ggtt_bind_put_ce(ce);
+	return true;
+
+err_rq:
+	i915_request_put(rq);
+put_ce:
+	gen8_ggtt_bind_put_ce(ce);
+	return false;
+}
+
 static void gen8_set_pte(void __iomem *addr, gen8_pte_t pte)
 {
 	writeq(pte, addr);
@@ -321,6 +466,21 @@ static void gen8_ggtt_insert_page(struct i915_address_space *vm,
 	ggtt->invalidate(ggtt);
 }
 
+static void gen8_ggtt_insert_page_bind(struct i915_address_space *vm,
+				       dma_addr_t addr, u64 offset,
+				       unsigned int pat_index, u32 flags)
+{
+	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm);
+	gen8_pte_t pte;
+
+	pte = ggtt->vm.pte_encode(addr, pat_index, flags);
+	if (should_update_ggtt_with_bind(i915_vm_to_ggtt(vm)) &&
+	    gen8_ggtt_bind_ptes(ggtt, offset, NULL, 1, pte))
+		return ggtt->invalidate(ggtt);
+
+	gen8_ggtt_insert_page(vm, addr, offset, pat_index, flags);
+}
+
 static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 				     struct i915_vma_resource *vma_res,
 				     unsigned int pat_index,
@@ -360,6 +520,50 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 	ggtt->invalidate(ggtt);
 }
 
+static bool __gen8_ggtt_insert_entries_bind(struct i915_address_space *vm,
+					    struct i915_vma_resource *vma_res,
+					    unsigned int pat_index, u32 flags)
+{
+	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm);
+	gen8_pte_t scratch_pte = vm->scratch[0]->encode;
+	gen8_pte_t pte_encode;
+	u64 start, end;
+
+	pte_encode = ggtt->vm.pte_encode(0, pat_index, flags);
+	start = (vma_res->start - vma_res->guard) / I915_GTT_PAGE_SIZE;
+	end = start + vma_res->guard / I915_GTT_PAGE_SIZE;
+	if (!gen8_ggtt_bind_ptes(ggtt, start, NULL, end - start, scratch_pte))
+		goto err;
+
+	start = end;
+	end += (vma_res->node_size + vma_res->guard) / I915_GTT_PAGE_SIZE;
+	if (!gen8_ggtt_bind_ptes(ggtt, start, vma_res->bi.pages,
+	      vma_res->node_size / I915_GTT_PAGE_SIZE, pte_encode))
+		goto err;
+
+	start += vma_res->node_size / I915_GTT_PAGE_SIZE;
+	if (!gen8_ggtt_bind_ptes(ggtt, start, NULL, end - start, scratch_pte))
+		goto err;
+
+	return true;
+
+err:
+	return false;
+}
+
+static void gen8_ggtt_insert_entries_bind(struct i915_address_space *vm,
+					  struct i915_vma_resource *vma_res,
+					  unsigned int pat_index, u32 flags)
+{
+	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm);
+
+	if (should_update_ggtt_with_bind(i915_vm_to_ggtt(vm)) &&
+	    __gen8_ggtt_insert_entries_bind(vm, vma_res, pat_index, flags))
+		return ggtt->invalidate(ggtt);
+
+	gen8_ggtt_insert_entries(vm, vma_res, pat_index, flags);
+}
+
 static void gen8_ggtt_clear_range(struct i915_address_space *vm,
 				  u64 start, u64 length)
 {
@@ -381,6 +585,27 @@ static void gen8_ggtt_clear_range(struct i915_address_space *vm,
 		gen8_set_pte(&gtt_base[i], scratch_pte);
 }
 
+static void gen8_ggtt_scratch_range_bind(struct i915_address_space *vm,
+					 u64 start, u64 length)
+{
+	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm);
+	unsigned int first_entry = start / I915_GTT_PAGE_SIZE;
+	unsigned int num_entries = length / I915_GTT_PAGE_SIZE;
+	const gen8_pte_t scratch_pte = vm->scratch[0]->encode;
+	const int max_entries = ggtt_total_entries(ggtt) - first_entry;
+
+	if (WARN(num_entries > max_entries,
+		 "First entry = %d; Num entries = %d (max=%d)\n",
+		 first_entry, num_entries, max_entries))
+		num_entries = max_entries;
+
+	if (should_update_ggtt_with_bind(ggtt) && gen8_ggtt_bind_ptes(ggtt, first_entry,
+	     NULL, num_entries, scratch_pte))
+		return ggtt->invalidate(ggtt);
+
+	gen8_ggtt_clear_range(vm, start, length);
+}
+
 static void gen6_ggtt_insert_page(struct i915_address_space *vm,
 				  dma_addr_t addr,
 				  u64 offset,
@@ -947,13 +1172,20 @@ static unsigned int gen6_gttadr_offset(struct drm_i915_private *i915)
 static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 {
 	struct drm_i915_private *i915 = ggtt->vm.i915;
+	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
 	struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
 	phys_addr_t phys_addr;
 	u32 pte_flags;
 	int ret;
 
 	GEM_WARN_ON(pci_resource_len(pdev, GEN4_GTTMMADR_BAR) != gen6_gttmmadr_size(i915));
-	phys_addr = pci_resource_start(pdev, GEN4_GTTMMADR_BAR) + gen6_gttadr_offset(i915);
+
+	if (i915_direct_stolen_access(i915)) {
+		drm_dbg(&i915->drm, "Using direct GSM access\n");
+		phys_addr = intel_uncore_read64(uncore, GEN6_GSMBASE) & GEN11_BDSM_MASK;
+	} else {
+		phys_addr = pci_resource_start(pdev, GEN4_GTTMMADR_BAR) + gen6_gttadr_offset(i915);
+	}
 
 	if (needs_wc_ggtt_mapping(i915))
 		ggtt->gsm = ioremap_wc(phys_addr, size);
@@ -992,6 +1224,7 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 {
 	struct drm_i915_private *i915 = ggtt->vm.i915;
+	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
 	struct pci_dev *pdev = i915->drm.pdev;
 	phys_addr_t phys_addr;
 	bus_addr_t addr;
@@ -1008,7 +1241,13 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 		return ret;
 
 	GEM_WARN_ON(len != gen6_gttmmadr_size(i915));
-	phys_addr = addr + gen6_gttadr_offset(i915);
+
+	if (i915_direct_stolen_access(i915)) {
+		drm_dbg(&i915->drm, "Using direct GSM access\n");
+		phys_addr = intel_uncore_read64(uncore, GEN6_GSMBASE) & GEN11_BDSM_MASK;
+	} else {
+		phys_addr = addr + gen6_gttadr_offset(i915);
+	}
 
 	if (needs_wc_ggtt_mapping(i915))
 		flags = BUS_SPACE_MAP_PREFETCHABLE;
@@ -1141,6 +1380,17 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 			I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND;
 	}
 
+	if (i915_ggtt_require_binder(i915)) {
+		ggtt->vm.scratch_range = gen8_ggtt_scratch_range_bind;
+		ggtt->vm.insert_page = gen8_ggtt_insert_page_bind;
+		ggtt->vm.insert_entries = gen8_ggtt_insert_entries_bind;
+		/*
+		 * On GPU is hung, we might bind VMAs for error capture.
+		 * Fallback to CPU GGTT updates in that case.
+		 */
+		ggtt->vm.raw_insert_page = gen8_ggtt_insert_page;
+	}
+
 	if (intel_uc_wants_guc(&ggtt->vm.gt->uc))
 		ggtt->invalidate = guc_ggtt_invalidate;
 	else

sys/dev/pci/drm/i915/gt/intel_gt.c

@@ -1073,6 +1073,55 @@ err:
 
 #endif
 
+static void __intel_gt_bind_context_set_ready(struct intel_gt *gt, bool ready)
+{
+	struct intel_engine_cs *engine = gt->engine[BCS0];
+
+	if (engine && engine->bind_context)
+		engine->bind_context_ready = ready;
+}
+
+/**
+ * intel_gt_bind_context_set_ready - Set the context binding as ready
+ *
+ * @gt: GT structure
+ *
+ * This function marks the binder context as ready.
+ */
+void intel_gt_bind_context_set_ready(struct intel_gt *gt)
+{
+	__intel_gt_bind_context_set_ready(gt, true);
+}
+
+/**
+ * intel_gt_bind_context_set_unready - Set the context binding as ready
+ * @gt: GT structure
+ *
+ * This function marks the binder context as not ready.
+ */
+
+void intel_gt_bind_context_set_unready(struct intel_gt *gt)
+{
+	__intel_gt_bind_context_set_ready(gt, false);
+}
+
+/**
+ * intel_gt_is_bind_context_ready - Check if context binding is ready
+ *
+ * @gt: GT structure
+ *
+ * This function returns binder context's ready status.
+ */
+bool intel_gt_is_bind_context_ready(struct intel_gt *gt)
+{
+	struct intel_engine_cs *engine = gt->engine[BCS0];
+
+	if (engine)
+		return engine->bind_context_ready;
+
+	return false;
+}
+
 int intel_gt_tiles_init(struct drm_i915_private *i915)
 {
 	struct intel_gt *gt;

sys/dev/pci/drm/i915/gt/intel_gt.h

@@ -148,4 +148,7 @@ enum i915_map_type intel_gt_coherent_map_type(struct intel_gt *gt,
 					      struct drm_i915_gem_object *obj,
 					      bool always_coherent);
 
+void intel_gt_bind_context_set_ready(struct intel_gt *gt);
+void intel_gt_bind_context_set_unready(struct intel_gt *gt);
+bool intel_gt_is_bind_context_ready(struct intel_gt *gt);
 #endif /* __INTEL_GT_H__ */

sys/dev/pci/drm/i915/gt/intel_gt_pm.c

@@ -296,6 +296,7 @@ int intel_gt_resume(struct intel_gt *gt)
 out_fw:
 	intel_uncore_forcewake_put(gt->uncore, FORCEWAKE_ALL);
 	intel_gt_pm_put(gt);
+	intel_gt_bind_context_set_ready(gt);
 	return err;
 
 err_wedged:
@@ -322,6 +323,7 @@ static void wait_for_suspend(struct intel_gt *gt)
 
 void intel_gt_suspend_prepare(struct intel_gt *gt)
 {
+	intel_gt_bind_context_set_unready(gt);
 	user_forcewake(gt, true);
 	wait_for_suspend(gt);
 }
@@ -375,6 +377,7 @@ void intel_gt_suspend_late(struct intel_gt *gt)
 
 void intel_gt_runtime_suspend(struct intel_gt *gt)
 {
+	intel_gt_bind_context_set_unready(gt);
 	intel_uc_runtime_suspend(&gt->uc);
 
 	GT_TRACE(gt, "\n");
@@ -392,6 +395,7 @@ int intel_gt_runtime_resume(struct intel_gt *gt)
 	if (ret)
 		return ret;
 
+	intel_gt_bind_context_set_ready(gt);
 	return 0;
 }
 

sys/dev/pci/drm/i915/gt/intel_gtt.c

@@ -21,6 +21,12 @@
 #include "intel_gt_regs.h"
 #include "intel_gtt.h"
 
+bool i915_ggtt_require_binder(struct drm_i915_private *i915)
+{
+	/* Wa_13010847436 & Wa_14019519902 */
+	return !i915_direct_stolen_access(i915) &&
+		MEDIA_VER_FULL(i915) == IP_VER(13, 0);
+}
 
 static bool intel_ggtt_update_needs_vtd_wa(struct drm_i915_private *i915)
 {

sys/dev/pci/drm/i915/gt/intel_gtt.h

@@ -171,6 +171,9 @@ struct intel_gt;
 #define for_each_sgt_daddr(__dp, __iter, __sgt) \
 	__for_each_sgt_daddr(__dp, __iter, __sgt, I915_GTT_PAGE_SIZE)
 
+#define for_each_sgt_daddr_next(__dp, __iter) \
+	__for_each_daddr_next(__dp, __iter, I915_GTT_PAGE_SIZE)
+
 struct i915_page_table {
 	struct drm_i915_gem_object *base;
 	union {
@@ -690,4 +693,6 @@ static inline struct sgt_dma {
 	return (struct sgt_dma){ sg, addr, addr + sg_dma_len(sg) };
 }
 
+bool i915_ggtt_require_binder(struct drm_i915_private *i915);
+
 #endif

sys/dev/pci/drm/i915/gt/intel_region_lmem.c

@@ -157,8 +157,8 @@ region_lmem_init(struct intel_memory_region *mem)
 
 #ifdef __linux__
 	if (!io_mapping_init_wc(&mem->iomap,
-				mem->io_start,
-				mem->io_size))
+				mem->io.start,
+				resource_size(&mem->io)))
 		return -EIO;
 #else
 	struct drm_i915_private *i915 = mem->i915;
@@ -167,20 +167,20 @@ region_lmem_init(struct intel_memory_region *mem)
 	int i;
 	bus_space_handle_t bsh;
 
-	start = atop(mem->io_start);
-	end = start + atop(mem->io_size);
+	start = atop(mem->io.start);
+	end = start + atop(resource_size(&mem->io));
 	uvm_page_physload(start, end, start, end, PHYSLOAD_DEVICE);
 
-	pgs = PHYS_TO_VM_PAGE(mem->io_start);
-	for (i = 0; i < atop(mem->io_size); i++)
+	pgs = PHYS_TO_VM_PAGE(mem->io.start);
+	for (i = 0; i < atop(resource_size(&mem->io)); i++)
 		atomic_setbits_int(&(pgs[i].pg_flags), PG_PMAP_WC);
 
-	if (bus_space_map(i915->bst, mem->io_start, mem->io_size,
+	if (bus_space_map(i915->bst, mem->io.start, resource_size(&mem->io),
 	    BUS_SPACE_MAP_LINEAR | BUS_SPACE_MAP_PREFETCHABLE, &bsh))
 		panic("can't map lmem");
 
-	mem->iomap.base = mem->io_start;
-	mem->iomap.size = mem->io_size;
+	mem->iomap.base = mem->io.start;
+	mem->iomap.size = resource_size(&mem->io);
 	mem->iomap.iomem = bus_space_vaddr(i915->bst, bsh);
 #endif
 
@@ -280,7 +280,7 @@ static struct intel_memory_region *setup_lmem(struct intel_gt *gt)
 		lmem_size -= tile_stolen;
 	} else {
 		/* Stolen starts from GSMBASE without CCS */
-		lmem_size = intel_uncore_read64(&i915->uncore, GEN12_GSMBASE);
+		lmem_size = intel_uncore_read64(&i915->uncore, GEN6_GSMBASE);
 	}
 
 	i915_resize_lmem_bar(i915, lmem_size);
@@ -326,14 +326,6 @@ static struct intel_memory_region *setup_lmem(struct intel_gt *gt)
 	if (err)
 		goto err_region_put;
 
-	drm_dbg(&i915->drm, "Local memory: %pR\n", &mem->region);
-	drm_dbg(&i915->drm, "Local memory IO start: %pa\n",
-		&mem->io_start);
-	drm_info(&i915->drm, "Local memory IO size: %pa\n",
-		 &mem->io_size);
-	drm_info(&i915->drm, "Local memory available: %pa\n",
-		 &lmem_size);
-
 	if (io_size < lmem_size)
 		drm_info(&i915->drm, "Using a reduced BAR size of %lluMiB. Consider enabling 'Resizable BAR' or similar, if available in the BIOS.\n",
 			 (u64)io_size >> 20);

sys/dev/pci/drm/i915/gt/selftest_tlb.c

@@ -199,8 +199,8 @@ static struct drm_i915_gem_object *create_lmem(struct intel_gt *gt)
 	 * of pages. To succeed with both allocations, especially in case of Small
 	 * BAR, try to allocate no more than quarter of mappable memory.
 	 */
-	if (mr && size > mr->io_size / 4)
-		size = mr->io_size / 4;
+	if (mr && size > resource_size(&mr->io) / 4)
+		size = resource_size(&mr->io) / 4;
 
 	return i915_gem_object_create_lmem(gt->i915, size, I915_BO_ALLOC_CONTIGUOUS);
 }

sys/dev/pci/drm/i915/i915_devlist.h

@@ -321,4 +321,9 @@ static const struct pci_matchid i915_devices[] = {
 	{ 0x8086, 0xa7a8 },
 	{ 0x8086, 0xa7aa },
 	{ 0x8086, 0xa7ab },
+	{ 0x8086, 0x7d40 },
+	{ 0x8086, 0x7d60 },
+	{ 0x8086, 0x7d45 },
+	{ 0x8086, 0x7d55 },
+	{ 0x8086, 0x7dd5 },
 };

sys/dev/pci/drm/i915/i915_driver.c

@@ -2263,6 +2263,11 @@ inteldrm_attach(struct device *parent, struct device *self, void *aux)
 	dev_priv->memex = pa->pa_memex;
 	dev_priv->vga_regs = &dev_priv->bar;
 
+	id = drm_find_description(PCI_VENDOR(pa->pa_id),
+	    PCI_PRODUCT(pa->pa_id), pciidlist);
+	dev_priv->id = id;
+	info = (struct intel_device_info *)id->driver_data;
+
 	if (PCI_CLASS(pa->pa_class) == PCI_CLASS_DISPLAY &&
 	    PCI_SUBCLASS(pa->pa_class) == PCI_SUBCLASS_DISPLAY_VGA &&
 	    (pci_conf_read(pa->pa_pc, pa->pa_tag, PCI_COMMAND_STATUS_REG)
@@ -2281,6 +2286,18 @@ inteldrm_attach(struct device *parent, struct device *self, void *aux)
 	}
 #endif
 
+	/*
+	 * Meteor Lake GOP framebuffer doesn't pass efifb pci bar tests
+	 * too early for IS_METEORLAKE which uses runtime info
+	 */
+	if (info->platform == INTEL_METEORLAKE) {
+		dev_priv->primary = 1;
+		dev_priv->console = 1;
+#if NEFIFB > 0
+		efifb_detach();
+#endif
+	}
+
 	printf("\n");
 
 	dev = drm_attach_pci(&i915_drm_driver, pa, 0, dev_priv->primary,
@@ -2290,11 +2307,6 @@ inteldrm_attach(struct device *parent, struct device *self, void *aux)
 		return;
 	}
 
-	id = drm_find_description(PCI_VENDOR(pa->pa_id),
-	    PCI_PRODUCT(pa->pa_id), pciidlist);
-	dev_priv->id = id;
-	info = (struct intel_device_info *)id->driver_data;
-
 	/* Device parameters start as a copy of module parameters. */
 	i915_params_copy(&dev_priv->params, &i915_modparams);
 	dev_priv->params.request_timeout_ms = 0;

sys/dev/pci/drm/i915/i915_gpu_error.c

@@ -1206,7 +1206,7 @@ i915_vma_coredump_create(const struct intel_gt *gt,
 			dma_addr_t offset = dma - mem->region.start;
 			void __iomem *s;
 
-			if (offset + PAGE_SIZE > mem->io_size) {
+			if (offset + PAGE_SIZE > resource_size(&mem->io)) {
 				ret = -EINVAL;
 				break;
 			}

sys/dev/pci/drm/i915/i915_pci.c

@@ -838,7 +838,6 @@ static const struct intel_device_info mtl_info = {
 	.has_pxp = 1,
 	.memory_regions = REGION_SMEM | REGION_STOLEN_LMEM,
 	.platform_engine_mask = BIT(RCS0) | BIT(BCS0) | BIT(CCS0),
-	.require_force_probe = 1,
 	MTL_CACHELEVEL,
 };
 

sys/dev/pci/drm/i915/i915_perf.c

@@ -2803,26 +2803,6 @@ oa_configure_all_contexts(struct i915_perf_stream *stream,
 	return 0;
 }
 
-static int
-gen12_configure_all_contexts(struct i915_perf_stream *stream,
-			     const struct i915_oa_config *oa_config,
-			     struct i915_active *active)
-{
-	struct flex regs[] = {
-		{
-			GEN8_R_PWR_CLK_STATE(RENDER_RING_BASE),
-			CTX_R_PWR_CLK_STATE,
-		},
-	};
-
-	if (stream->engine->class != RENDER_CLASS)
-		return 0;
-
-	return oa_configure_all_contexts(stream,
-					 regs, ARRAY_SIZE(regs),
-					 active);
-}
-
 static int
 lrc_configure_all_contexts(struct i915_perf_stream *stream,
 			   const struct i915_oa_config *oa_config,
@@ -2929,7 +2909,6 @@ gen12_enable_metric_set(struct i915_perf_stream *stream,
 {
 	struct drm_i915_private *i915 = stream->perf->i915;
 	struct intel_uncore *uncore = stream->uncore;
-	struct i915_oa_config *oa_config = stream->oa_config;
 	bool periodic = stream->periodic;
 	u32 period_exponent = stream->period_exponent;
 	u32 sqcnt1;
@@ -2973,15 +2952,6 @@ gen12_enable_metric_set(struct i915_perf_stream *stream,
 
 	intel_uncore_rmw(uncore, GEN12_SQCNT1, 0, sqcnt1);
 
-	/*
-	 * Update all contexts prior writing the mux configurations as we need
-	 * to make sure all slices/subslices are ON before writing to NOA
-	 * registers.
-	 */
-	ret = gen12_configure_all_contexts(stream, oa_config, active);
-	if (ret)
-		return ret;
-
 	/*
 	 * For Gen12, performance counters are context
 	 * saved/restored. Only enable it for the context that
@@ -3036,9 +3006,6 @@ static void gen12_disable_metric_set(struct i915_perf_stream *stream)
 				   _MASKED_BIT_DISABLE(GEN12_DISABLE_DOP_GATING));
 	}
 
-	/* Reset all contexts' slices/subslices configurations. */
-	gen12_configure_all_contexts(stream, NULL, NULL);
-
 	/* disable the context save/restore or OAR counters */
 	if (stream->ctx)
 		gen12_configure_oar_context(stream, NULL);

sys/dev/pci/drm/i915/i915_query.c

@@ -513,7 +513,7 @@ static int query_memregion_info(struct drm_i915_private *i915,
 		info.probed_size = mr->total;
 
 		if (mr->type == INTEL_MEMORY_LOCAL)
-			info.probed_cpu_visible_size = mr->io_size;
+			info.probed_cpu_visible_size = resource_size(&mr->io);
 		else
 			info.probed_cpu_visible_size = mr->total;
 

sys/dev/pci/drm/i915/i915_reg.h

@@ -5452,6 +5452,9 @@
 #define   GEN6_PCODE_FREQ_RING_RATIO_SHIFT	16
 #define GEN6_PCODE_DATA1			_MMIO(0x13812C)
 
+#define MTL_PCODE_STOLEN_ACCESS			_MMIO(0x138914)
+#define   STOLEN_ACCESS_ALLOWED			0x1
+
 /* IVYBRIDGE DPF */
 #define GEN7_L3CDERRST1(slice)		_MMIO(0xB008 + (slice) * 0x200) /* L3CD Error Status 1 */
 #define   GEN7_L3CDERRST1_ROW_MASK	(0x7ff << 14)
@@ -6581,9 +6584,10 @@ enum skl_power_gate {
 #define   GMS_MASK			REG_GENMASK(15, 8)
 #define   GGMS_MASK			REG_GENMASK(7, 6)
 
-#define GEN12_GSMBASE			_MMIO(0x108100)
-#define GEN12_DSMBASE			_MMIO(0x1080C0)
-#define   GEN12_BDSM_MASK		REG_GENMASK64(63, 20)
+#define GEN6_GSMBASE			_MMIO(0x108100)
+#define GEN6_DSMBASE			_MMIO(0x1080C0)
+#define   GEN6_BDSM_MASK		REG_GENMASK64(31, 20)
+#define   GEN11_BDSM_MASK		REG_GENMASK64(63, 20)
 
 #define XEHP_CLOCK_GATE_DIS		_MMIO(0x101014)
 #define   SGSI_SIDECLK_DIS		REG_BIT(17)

sys/dev/pci/drm/i915/i915_scatterlist.h

@@ -91,6 +91,16 @@ static inline struct scatterlist *__sg_next(struct scatterlist *sg)
 	     ((__dp) = (__iter).dma + (__iter).curr), (__iter).sgp;	\
 	     (((__iter).curr += (__step)) >= (__iter).max) ?		\
 	     (__iter) = __sgt_iter(__sg_next((__iter).sgp), true), 0 : 0)
+/**
+ * __for_each_daddr_next - iterates over the device addresses with pre-initialized iterator.
+ * @__dp:	Device address (output)
+ * @__iter:	'struct sgt_iter' (iterator state, external)
+ * @__step:	step size
+ */
+#define __for_each_daddr_next(__dp, __iter, __step)                  \
+	for (; ((__dp) = (__iter).dma + (__iter).curr), (__iter).sgp;   \
+	     (((__iter).curr += (__step)) >= (__iter).max) ?            \
+	     (__iter) = __sgt_iter(__sg_next((__iter).sgp), true), 0 : 0)
 
 /**
  * for_each_sgt_page - iterate over the pages of the given sg_table

sys/dev/pci/drm/i915/i915_utils.c

@@ -8,6 +8,7 @@
 #include <drm/drm_drv.h>
 
 #include "i915_drv.h"
+#include "i915_reg.h"
 #include "i915_utils.h"
 
 #include <sys/syslog.h>
@@ -138,3 +139,19 @@ bool i915_vtd_active(struct drm_i915_private *i915)
 	return i915_run_as_guest();
 #endif
 }
+
+bool i915_direct_stolen_access(struct drm_i915_private *i915)
+{
+	/*
+	 * Wa_22018444074
+	 *
+	 * Access via BAR can hang MTL, go directly to GSM/DSM,
+	 * except for VM guests which won't have access to it.
+	 *
+	 * Normally this would not work but on MTL the system firmware
+	 * should have relaxed the access permissions sufficiently.
+	 * 0x138914==0x1 indicates that the firmware has done its job.
+	 */
+	return IS_METEORLAKE(i915) && !i915_run_as_guest() &&
+		intel_uncore_read(&i915->uncore, MTL_PCODE_STOLEN_ACCESS) == STOLEN_ACCESS_ALLOWED;
+}

sys/dev/pci/drm/i915/i915_utils.h

@@ -401,4 +401,6 @@ static inline bool i915_run_as_guest(void)
 
 bool i915_vtd_active(struct drm_i915_private *i915);
 
+bool i915_direct_stolen_access(struct drm_i915_private *i915);
+
 #endif /* !__I915_UTILS_H */

sys/dev/pci/drm/i915/intel_memory_region.c

@@ -50,7 +50,7 @@ static int __iopagetest(struct intel_memory_region *mem,
 	if (memchr_inv(result, value, sizeof(result))) {
 		dev_err(mem->i915->drm.dev,
 			"Failed to read back from memory region:%pR at [%pa + %pa] for %ps; wrote %x, read (%x, %x, %x)\n",
-			&mem->region, &mem->io_start, &offset, caller,
+			&mem->region, &mem->io.start, &offset, caller,
 			value, result[0], result[1], result[2]);
 		return -EINVAL;
 	}
@@ -70,11 +70,11 @@ static int iopagetest(struct intel_memory_region *mem,
 	int err;
 	int i;
 
-	va = ioremap_wc(mem->io_start + offset, PAGE_SIZE);
+	va = ioremap_wc(mem->io.start + offset, PAGE_SIZE);
 	if (!va) {
 		dev_err(mem->i915->drm.dev,
 			"Failed to ioremap memory region [%pa + %pa] for %ps\n",
-			&mem->io_start, &offset, caller);
+			&mem->io.start, &offset, caller);
 		return -EFAULT;
 	}
 
@@ -106,10 +106,10 @@ static int iomemtest(struct intel_memory_region *mem,
 	resource_size_t last, page;
 	int err;
 
-	if (mem->io_size < PAGE_SIZE)
+	if (resource_size(&mem->io) < PAGE_SIZE)
 		return 0;
 
-	last = mem->io_size - PAGE_SIZE;
+	last = resource_size(&mem->io) - PAGE_SIZE;
 
 	/*
 	 * Quick test to check read/write access to the iomap (backing store).
@@ -211,7 +211,7 @@ static int intel_memory_region_memtest(struct intel_memory_region *mem,
 	struct drm_i915_private *i915 = mem->i915;
 	int err = 0;
 
-	if (!mem->io_start)
+	if (!mem->io.start)
 		return 0;
 
 	if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM) || i915->params.memtest)
@@ -240,8 +240,7 @@ intel_memory_region_create(struct drm_i915_private *i915,
 
 	mem->i915 = i915;
 	mem->region = DEFINE_RES_MEM(start, size);
-	mem->io_start = io_start;
-	mem->io_size = io_size;
+	mem->io = DEFINE_RES_MEM(io_start, io_size);
 	mem->min_page_size = min_page_size;
 	mem->ops = ops;
 	mem->total = size;
@@ -360,6 +359,24 @@ int intel_memory_regions_hw_probe(struct drm_i915_private *i915)
 		i915->mm.regions[i] = mem;
 	}
 
+	for (i = 0; i < ARRAY_SIZE(i915->mm.regions); i++) {
+		struct intel_memory_region *mem = i915->mm.regions[i];
+		u64 region_size, io_size;
+
+		if (!mem)
+			continue;
+
+		region_size = resource_size(&mem->region) >> 20;
+		io_size = resource_size(&mem->io) >> 20;
+
+		if (resource_size(&mem->io))
+			drm_dbg(&i915->drm, "Memory region(%d): %s: %llu MiB %pR, io: %llu MiB %pR\n",
+				mem->id, mem->name, region_size, &mem->region, io_size, &mem->io);
+		else
+			drm_dbg(&i915->drm, "Memory region(%d): %s: %llu MiB %pR, io: n/a\n",
+				mem->id, mem->name, region_size, &mem->region);
+	}
+
 	return 0;
 
 out_cleanup:

sys/dev/pci/drm/i915/intel_memory_region.h

@@ -73,8 +73,7 @@ struct intel_memory_region {
 	struct io_mapping iomap;
 	struct resource region;
 
-	resource_size_t io_start;
-	resource_size_t io_size;
+	struct resource io;
 	resource_size_t min_page_size;
 	resource_size_t total;
 

sys/dev/pci/drm/i915/intel_region_ttm.c

@@ -93,7 +93,7 @@ int intel_region_ttm_init(struct intel_memory_region *mem)
 
 	ret = i915_ttm_buddy_man_init(bdev, mem_type, false,
 				      resource_size(&mem->region),
-				      mem->io_size,
+				      resource_size(&mem->io),
 				      mem->min_page_size, PAGE_SIZE);
 	if (ret)
 		return ret;
@@ -225,16 +225,16 @@ intel_region_ttm_resource_alloc(struct intel_memory_region *mem,
 			goto out;
 		}
 		place.lpfn = place.fpfn + (size >> PAGE_SHIFT);
-	} else if (mem->io_size && mem->io_size < mem->total) {
+	} else if (resource_size(&mem->io) && resource_size(&mem->io) < mem->total) {
 		if (flags & I915_BO_ALLOC_GPU_ONLY) {
 			place.flags |= TTM_PL_FLAG_TOPDOWN;
 		} else {
 			place.fpfn = 0;
-			if (WARN_ON(overflows_type(mem->io_size >> PAGE_SHIFT, place.lpfn))) {
+			if (WARN_ON(overflows_type(resource_size(&mem->io) >> PAGE_SHIFT, place.lpfn))) {
 				ret = -E2BIG;
 				goto out;
 			}
-			place.lpfn = mem->io_size >> PAGE_SHIFT;
+			place.lpfn = resource_size(&mem->io) >> PAGE_SHIFT;
 		}
 	}
 

sys/dev/pci/drm/i915/intel_uncore.c

@@ -1833,7 +1833,10 @@ static const struct intel_forcewake_range __mtl_fw_ranges[] = {
 	GEN_FW_RANGE(0x24000, 0x2ffff, 0), /*
 		0x24000 - 0x2407f: always on
 		0x24080 - 0x2ffff: reserved */
-	GEN_FW_RANGE(0x30000, 0x3ffff, FORCEWAKE_GT)
+	GEN_FW_RANGE(0x30000, 0x3ffff, FORCEWAKE_GT),
+	GEN_FW_RANGE(0x40000, 0x1901ef, 0),
+	GEN_FW_RANGE(0x1901f0, 0x1901f3, FORCEWAKE_GT)
+		/* FIXME: WA to wake GT while triggering H2G */
 };
 
 /*

sys/dev/pci/drm/i915/intel_wakeref.c

@@ -10,21 +10,12 @@
 #include "intel_wakeref.h"
 #include "i915_drv.h"
 
-static void rpm_get(struct intel_wakeref *wf)
-{
-	wf->wakeref = intel_runtime_pm_get(&wf->i915->runtime_pm);
-}
-
-static void rpm_put(struct intel_wakeref *wf)
-{
-	intel_wakeref_t wakeref = fetch_and_zero(&wf->wakeref);
-
-	intel_runtime_pm_put(&wf->i915->runtime_pm, wakeref);
-	INTEL_WAKEREF_BUG_ON(!wakeref);
-}
-
 int __intel_wakeref_get_first(struct intel_wakeref *wf)
 {
+	intel_wakeref_t wakeref;
+	int ret = 0;
+
+	wakeref = intel_runtime_pm_get(&wf->i915->runtime_pm);
 	/*
 	 * Treat get/put as different subclasses, as we may need to run
 	 * the put callback from under the shrinker and do not want to
@@ -32,41 +23,52 @@ int __intel_wakeref_get_first(struct intel_wakeref *wf)
 	 * upon acquiring the wakeref.
 	 */
 	mutex_lock_nested(&wf->mutex, SINGLE_DEPTH_NESTING);
+
 	if (!atomic_read(&wf->count)) {
-		int err;
+		INTEL_WAKEREF_BUG_ON(wf->wakeref);
+		wf->wakeref = wakeref;
+		wakeref = 0;
 
-		rpm_get(wf);
-
-		err = wf->ops->get(wf);
-		if (unlikely(err)) {
-			rpm_put(wf);
-			mutex_unlock(&wf->mutex);
-			return err;
+		ret = wf->ops->get(wf);
+		if (ret) {
+			wakeref = xchg(&wf->wakeref, 0);
+			wake_up_var(&wf->wakeref);
+			goto unlock;
 		}
 
 		smp_mb__before_atomic(); /* release wf->count */
 	}
-	atomic_inc(&wf->count);
-	mutex_unlock(&wf->mutex);
 
+	atomic_inc(&wf->count);
 	INTEL_WAKEREF_BUG_ON(atomic_read(&wf->count) <= 0);
-	return 0;
+
+unlock:
+	mutex_unlock(&wf->mutex);
+	if (unlikely(wakeref))
+		intel_runtime_pm_put(&wf->i915->runtime_pm, wakeref);
+
+	return ret;
 }
 
 static void ____intel_wakeref_put_last(struct intel_wakeref *wf)
 {
+	intel_wakeref_t wakeref = 0;
+
 	INTEL_WAKEREF_BUG_ON(atomic_read(&wf->count) <= 0);
 	if (unlikely(!atomic_dec_and_test(&wf->count)))
 		goto unlock;
 
 	/* ops->put() must reschedule its own release on error/deferral */
 	if (likely(!wf->ops->put(wf))) {
-		rpm_put(wf);
+		INTEL_WAKEREF_BUG_ON(!wf->wakeref);
+		wakeref = xchg(&wf->wakeref, 0);
 		wake_up_var(&wf->wakeref);
 	}
 
 unlock:
 	mutex_unlock(&wf->mutex);
+	if (wakeref)
+		intel_runtime_pm_put(&wf->i915->runtime_pm, wakeref);
 }
 
 void __intel_wakeref_put_last(struct intel_wakeref *wf, unsigned long flags)

sys/dev/pci/drm/i915/pxp/intel_pxp_cmd_interface_43.h

@@ -14,8 +14,8 @@
 #define PXP43_CMDID_NEW_HUC_AUTH 0x0000003F /* MTL+ */
 #define PXP43_CMDID_INIT_SESSION 0x00000036
 
-/* PXP-Packet sizes for MTL's GSCCS-HECI instruction */
-#define PXP43_MAX_HECI_INOUT_SIZE (SZ_32K)
+/* PXP-Packet sizes for MTL's GSCCS-HECI instruction is spec'd at 65K before page alignment*/
+#define PXP43_MAX_HECI_INOUT_SIZE (PAGE_ALIGN(SZ_64K + SZ_1K))
 
 /* PXP-Packet size for MTL's NEW_HUC_AUTH instruction */
 #define PXP43_HUC_AUTH_INOUT_SIZE (SZ_4K)

sys/dev/pci/drm/i915/selftests/intel_memory_region.c

@@ -544,8 +544,8 @@ static u64 igt_object_mappable_total(struct drm_i915_gem_object *obj)
 		u64 start = drm_buddy_block_offset(block);
 		u64 end = start + drm_buddy_block_size(mm, block);
 
-		if (start < mr->io_size)
-			total += min_t(u64, end, mr->io_size) - start;
+		if (start < resource_size(&mr->io))
+			total += min_t(u64, end, resource_size(&mr->io)) - start;
 	}
 
 	return total;