Merge tag 'drm-fixes-2022-07-15' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "This is the regular fixes pull for this week. This has a bunch of
  amdgpu fixes, major one reverts the buddy allocator until it can be
  tested more, otherwise just small ones, then i915 has a bunch of
  fixes.

  The outstanding firmware regressions reported by phoronix will
  hopefully be dealt with ASAP.

  amdgpu:
   - revert buddy allocator support for now
   - DP MST blank screen fix for specific platforms
   - MEC firmware check fix for GC 10.3.7
   - Deep color fix for DCE
   - Fix possible divide by 0
   - Coverage blend mode fix
   - Fix cursor only commit timestamps

  i915:
   - Selftest fix
   - TTM fix sg_table construction
   - Error return fixes
   - Fix a performance regression related to waitboost
   - Fix GT resets"

* tag 'drm-fixes-2022-07-15' of git://anongit.freedesktop.org/drm/drm:
  drm/amd/display: Ensure valid event timestamp for cursor-only commits
  drm/amd/display: correct check of coverage blend mode
  drm/amd/pm: Prevent divide by zero
  drm/amd/display: Only use depth 36 bpp linebuffers on DCN display engines.
  drm/amdkfd: correct the MEC atomic support firmware checking for GC 10.3.7
  drm/amd/display: Ignore First MST Sideband Message Return Error
  drm/i915/selftests: fix subtraction overflow bug
  drm/i915/gem: Look for waitboosting across the whole object prior to individual waits
  drm/i915/gt: Serialize TLB invalidates with GT resets
  drm/i915/gt: Serialize GRDOM access between multiple engine resets
  drm/i915/ttm: fix sg_table construction
  drm/i915/selftests: fix a couple IS_ERR() vs NULL tests
  drm/i915: Fix vm use-after-free in vma destruction
  drm/i915/guc: ADL-N should use the same GuC FW as ADL-S
  drm/i915: fix a possible refcount leak in intel_dp_add_mst_connector()
  drm/i915/gvt: IS_ERR() vs NULL bug in intel_gvt_update_reg_whitelist()
  Revert "drm/amdgpu: add drm buddy support to amdgpu"

drivers/gpu/drm/Kconfig

@@ -256,7 +256,6 @@ config DRM_AMDGPU
 	select HWMON
 	select BACKLIGHT_CLASS_DEVICE
 	select INTERVAL_TREE
-	select DRM_BUDDY
 	help
 	  Choose this option if you have a recent AMD Radeon graphics card.
 

drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h

@@ -30,15 +30,12 @@
 #include <drm/ttm/ttm_resource.h>
 #include <drm/ttm/ttm_range_manager.h>
 
-#include "amdgpu_vram_mgr.h"
-
 /* state back for walking over vram_mgr and gtt_mgr allocations */
 struct amdgpu_res_cursor {
 	uint64_t		start;
 	uint64_t		size;
 	uint64_t		remaining;
-	void			*node;
-	uint32_t		mem_type;
+	struct drm_mm_node	*node;
 };
 
 /**
@@ -55,41 +52,19 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
 				    uint64_t start, uint64_t size,
 				    struct amdgpu_res_cursor *cur)
 {
-	struct drm_buddy_block *block;
-	struct list_head *head, *next;
 	struct drm_mm_node *node;
 
-	if (!res)
-		goto fallback;
+	if (!res || res->mem_type == TTM_PL_SYSTEM) {
+		cur->start = start;
+		cur->size = size;
+		cur->remaining = size;
+		cur->node = NULL;
+		WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
+		return;
+	}
 
 	BUG_ON(start + size > res->num_pages << PAGE_SHIFT);
 
-	cur->mem_type = res->mem_type;
-
-	switch (cur->mem_type) {
-	case TTM_PL_VRAM:
-		head = &to_amdgpu_vram_mgr_resource(res)->blocks;
-
-		block = list_first_entry_or_null(head,
-						 struct drm_buddy_block,
-						 link);
-		if (!block)
-			goto fallback;
-
-		while (start >= amdgpu_vram_mgr_block_size(block)) {
-			start -= amdgpu_vram_mgr_block_size(block);
-
-			next = block->link.next;
-			if (next != head)
-				block = list_entry(next, struct drm_buddy_block, link);
-		}
-
-		cur->start = amdgpu_vram_mgr_block_start(block) + start;
-		cur->size = min(amdgpu_vram_mgr_block_size(block) - start, size);
-		cur->remaining = size;
-		cur->node = block;
-		break;
-	case TTM_PL_TT:
 	node = to_ttm_range_mgr_node(res)->mm_nodes;
 	while (start >= node->size << PAGE_SHIFT)
 		start -= node++->size << PAGE_SHIFT;
@@ -98,20 +73,6 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
 	cur->size = min((node->size << PAGE_SHIFT) - start, size);
 	cur->remaining = size;
 	cur->node = node;
-		break;
-	default:
-		goto fallback;
-	}
-
-	return;
-
-fallback:
-	cur->start = start;
-	cur->size = size;
-	cur->remaining = size;
-	cur->node = NULL;
-	WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT);
-	return;
 }
 
 /**
@@ -124,9 +85,7 @@ static inline void amdgpu_res_first(struct ttm_resource *res,
  */
 static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
 {
-	struct drm_buddy_block *block;
-	struct drm_mm_node *node;
-	struct list_head *next;
+	struct drm_mm_node *node = cur->node;
 
 	BUG_ON(size > cur->remaining);
 
@@ -140,27 +99,9 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size)
 		return;
 	}
 
-	switch (cur->mem_type) {
-	case TTM_PL_VRAM:
-		block = cur->node;
-
-		next = block->link.next;
-		block = list_entry(next, struct drm_buddy_block, link);
-
-		cur->node = block;
-		cur->start = amdgpu_vram_mgr_block_start(block);
-		cur->size = min(amdgpu_vram_mgr_block_size(block), cur->remaining);
-		break;
-	case TTM_PL_TT:
-		node = cur->node;
-
 	cur->node = ++node;
 	cur->start = node->start << PAGE_SHIFT;
 	cur->size = min(node->size << PAGE_SHIFT, cur->remaining);
-		break;
-	default:
-		return;
-	}
 }
 
 #endif

drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h

@@ -26,7 +26,6 @@
 
 #include <linux/dma-direction.h>
 #include <drm/gpu_scheduler.h>
-#include "amdgpu_vram_mgr.h"
 #include "amdgpu.h"
 
 #define AMDGPU_PL_GDS		(TTM_PL_PRIV + 0)
@@ -39,6 +38,15 @@
 
 #define AMDGPU_POISON	0xd0bed0be
 
+struct amdgpu_vram_mgr {
+	struct ttm_resource_manager manager;
+	struct drm_mm mm;
+	spinlock_t lock;
+	struct list_head reservations_pending;
+	struct list_head reserved_pages;
+	atomic64_t vis_usage;
+};
+
 struct amdgpu_gtt_mgr {
 	struct ttm_resource_manager manager;
 	struct drm_mm mm;

drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h

-/* SPDX-License-Identifier: MIT
- * Copyright 2021 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-
-#ifndef __AMDGPU_VRAM_MGR_H__
-#define __AMDGPU_VRAM_MGR_H__
-
-#include <drm/drm_buddy.h>
-
-struct amdgpu_vram_mgr {
-	struct ttm_resource_manager manager;
-	struct drm_buddy mm;
-	/* protects access to buffer objects */
-	struct mutex lock;
-	struct list_head reservations_pending;
-	struct list_head reserved_pages;
-	atomic64_t vis_usage;
-	u64 default_page_size;
-};
-
-struct amdgpu_vram_mgr_resource {
-	struct ttm_resource base;
-	struct list_head blocks;
-	unsigned long flags;
-};
-
-static inline u64 amdgpu_vram_mgr_block_start(struct drm_buddy_block *block)
-{
-	return drm_buddy_block_offset(block);
-}
-
-static inline u64 amdgpu_vram_mgr_block_size(struct drm_buddy_block *block)
-{
-	return PAGE_SIZE << drm_buddy_block_order(block);
-}
-
-static inline struct drm_buddy_block *
-amdgpu_vram_mgr_first_block(struct list_head *list)
-{
-	return list_first_entry_or_null(list, struct drm_buddy_block, link);
-}
-
-static inline bool amdgpu_is_vram_mgr_blocks_contiguous(struct list_head *head)
-{
-	struct drm_buddy_block *block;
-	u64 start, size;
-
-	block = amdgpu_vram_mgr_first_block(head);
-	if (!block)
-		return false;
-
-	while (head != block->link.next) {
-		start = amdgpu_vram_mgr_block_start(block);
-		size = amdgpu_vram_mgr_block_size(block);
-
-		block = list_entry(block->link.next, struct drm_buddy_block, link);
-		if (start + size != amdgpu_vram_mgr_block_start(block))
-			return false;
-	}
-
-	return true;
-}
-
-static inline struct amdgpu_vram_mgr_resource *
-to_amdgpu_vram_mgr_resource(struct ttm_resource *res)
-{
-	return container_of(res, struct amdgpu_vram_mgr_resource, base);
-}
-
-#endif

drivers/gpu/drm/amd/amdkfd/kfd_device.c

@@ -184,6 +184,8 @@ static void kfd_device_info_init(struct kfd_dev *kfd,
 			/* Navi2x+, Navi1x+ */
 			if (gc_version == IP_VERSION(10, 3, 6))
 				kfd->device_info.no_atomic_fw_version = 14;
+			else if (gc_version == IP_VERSION(10, 3, 7))
+				kfd->device_info.no_atomic_fw_version = 3;
 			else if (gc_version >= IP_VERSION(10, 3, 0))
 				kfd->device_info.no_atomic_fw_version = 92;
 			else if (gc_version >= IP_VERSION(10, 1, 1))

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

@@ -72,6 +72,7 @@
 #include <linux/pci.h>
 #include <linux/firmware.h>
 #include <linux/component.h>
+#include <linux/dmi.h>
 
 #include <drm/display/drm_dp_mst_helper.h>
 #include <drm/display/drm_hdmi_helper.h>
@@ -462,6 +463,26 @@ static void dm_pflip_high_irq(void *interrupt_params)
 		     vrr_active, (int) !e);
 }
 
+static void dm_crtc_handle_vblank(struct amdgpu_crtc *acrtc)
+{
+	struct drm_crtc *crtc = &acrtc->base;
+	struct drm_device *dev = crtc->dev;
+	unsigned long flags;
+
+	drm_crtc_handle_vblank(crtc);
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+
+	/* Send completion event for cursor-only commits */
+	if (acrtc->event && acrtc->pflip_status != AMDGPU_FLIP_SUBMITTED) {
+		drm_crtc_send_vblank_event(crtc, acrtc->event);
+		drm_crtc_vblank_put(crtc);
+		acrtc->event = NULL;
+	}
+
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+}
+
 static void dm_vupdate_high_irq(void *interrupt_params)
 {
 	struct common_irq_params *irq_params = interrupt_params;
@@ -500,7 +521,7 @@ static void dm_vupdate_high_irq(void *interrupt_params)
 		 * if a pageflip happened inside front-porch.
 		 */
 		if (vrr_active) {
-			drm_crtc_handle_vblank(&acrtc->base);
+			dm_crtc_handle_vblank(acrtc);
 
 			/* BTR processing for pre-DCE12 ASICs */
 			if (acrtc->dm_irq_params.stream &&
@@ -552,7 +573,7 @@ static void dm_crtc_high_irq(void *interrupt_params)
 	 * to dm_vupdate_high_irq after end of front-porch.
 	 */
 	if (!vrr_active)
-		drm_crtc_handle_vblank(&acrtc->base);
+		dm_crtc_handle_vblank(acrtc);
 
 	/**
 	 * Following stuff must happen at start of vblank, for crc
@@ -1382,6 +1403,41 @@ static bool dm_should_disable_stutter(struct pci_dev *pdev)
 	return false;
 }
 
+static const struct dmi_system_id hpd_disconnect_quirk_table[] = {
+	{
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3660"),
+		},
+	},
+	{
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3260"),
+		},
+	},
+	{
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3460"),
+		},
+	},
+	{}
+};
+
+static void retrieve_dmi_info(struct amdgpu_display_manager *dm)
+{
+	const struct dmi_system_id *dmi_id;
+
+	dm->aux_hpd_discon_quirk = false;
+
+	dmi_id = dmi_first_match(hpd_disconnect_quirk_table);
+	if (dmi_id) {
+		dm->aux_hpd_discon_quirk = true;
+		DRM_INFO("aux_hpd_discon_quirk attached\n");
+	}
+}
+
 static int amdgpu_dm_init(struct amdgpu_device *adev)
 {
 	struct dc_init_data init_data;
@@ -1508,6 +1564,9 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 	}
 
 	INIT_LIST_HEAD(&adev->dm.da_list);
+
+	retrieve_dmi_info(&adev->dm);
+
 	/* Display Core create. */
 	adev->dm.dc = dc_create(&init_data);
 
@@ -5407,7 +5466,7 @@ fill_blending_from_plane_state(const struct drm_plane_state *plane_state,
 			}
 		}
 
-		if (per_pixel_alpha && plane_state->pixel_blend_mode == DRM_MODE_BLEND_COVERAGE)
+		if (*per_pixel_alpha && plane_state->pixel_blend_mode == DRM_MODE_BLEND_COVERAGE)
 			*pre_multiplied_alpha = false;
 	}
 
@@ -9135,6 +9194,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 	struct amdgpu_bo *abo;
 	uint32_t target_vblank, last_flip_vblank;
 	bool vrr_active = amdgpu_dm_vrr_active(acrtc_state);
+	bool cursor_update = false;
 	bool pflip_present = false;
 	struct {
 		struct dc_surface_update surface_updates[MAX_SURFACES];
@@ -9170,8 +9230,13 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 		struct dm_plane_state *dm_new_plane_state = to_dm_plane_state(new_plane_state);
 
 		/* Cursor plane is handled after stream updates */
-		if (plane->type == DRM_PLANE_TYPE_CURSOR)
+		if (plane->type == DRM_PLANE_TYPE_CURSOR) {
+			if ((fb && crtc == pcrtc) ||
+			    (old_plane_state->fb && old_plane_state->crtc == pcrtc))
+				cursor_update = true;
+
 			continue;
+		}
 
 		if (!fb || !crtc || pcrtc != crtc)
 			continue;
@@ -9334,6 +9399,17 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 				bundle->stream_update.vrr_infopacket =
 					&acrtc_state->stream->vrr_infopacket;
 		}
+	} else if (cursor_update && acrtc_state->active_planes > 0 &&
+		   !acrtc_state->force_dpms_off &&
+		   acrtc_attach->base.state->event) {
+		drm_crtc_vblank_get(pcrtc);
+
+		spin_lock_irqsave(&pcrtc->dev->event_lock, flags);
+
+		acrtc_attach->event = acrtc_attach->base.state->event;
+		acrtc_attach->base.state->event = NULL;
+
+		spin_unlock_irqrestore(&pcrtc->dev->event_lock, flags);
 	}
 
 	/* Update the planes if changed or disable if we don't have any. */

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

@@ -540,6 +540,14 @@ struct amdgpu_display_manager {
 	 * last successfully applied backlight values.
 	 */
 	u32 actual_brightness[AMDGPU_DM_MAX_NUM_EDP];
+
+	/**
+	 * @aux_hpd_discon_quirk:
+	 *
+	 * quirk for hpd discon while aux is on-going.
+	 * occurred on certain intel platform
+	 */
+	bool aux_hpd_discon_quirk;
 };
 
 enum dsc_clock_force_state {

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

@@ -56,6 +56,8 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 	ssize_t result = 0;
 	struct aux_payload payload;
 	enum aux_return_code_type operation_result;
+	struct amdgpu_device *adev;
+	struct ddc_service *ddc;
 
 	if (WARN_ON(msg->size > 16))
 		return -E2BIG;
@@ -74,6 +76,21 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 	result = dc_link_aux_transfer_raw(TO_DM_AUX(aux)->ddc_service, &payload,
 				      &operation_result);
 
+	/*
+	 * w/a on certain intel platform where hpd is unexpected to pull low during
+	 * 1st sideband message transaction by return AUX_RET_ERROR_HPD_DISCON
+	 * aux transaction is succuess in such case, therefore bypass the error
+	 */
+	ddc = TO_DM_AUX(aux)->ddc_service;
+	adev = ddc->ctx->driver_context;
+	if (adev->dm.aux_hpd_discon_quirk) {
+		if (msg->address == DP_SIDEBAND_MSG_DOWN_REQ_BASE &&
+			operation_result == AUX_RET_ERROR_HPD_DISCON) {
+			result = 0;
+			operation_result = AUX_RET_SUCCESS;
+		}
+	}
+
 	if (payload.write && result >= 0)
 		result = msg->size;
 

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

@@ -1117,12 +1117,13 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 	 * on certain displays, such as the Sharp 4k. 36bpp is needed
 	 * to support SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616 and
 	 * SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616 with actual > 10 bpc
-	 * precision on at least DCN display engines. However, at least
-	 * Carrizo with DCE_VERSION_11_0 does not like 36 bpp lb depth,
-	 * so use only 30 bpp on DCE_VERSION_11_0. Testing with DCE 11.2 and 8.3
-	 * did not show such problems, so this seems to be the exception.
+	 * precision on DCN display engines, but apparently not for DCE, as
+	 * far as testing on DCE-11.2 and DCE-8 showed. Various DCE parts have
+	 * problems: Carrizo with DCE_VERSION_11_0 does not like 36 bpp lb depth,
+	 * neither do DCE-8 at 4k resolution, or DCE-11.2 (broken identify pixel
+	 * passthrough). Therefore only use 36 bpp on DCN where it is actually needed.
 	 */
-	if (plane_state->ctx->dce_version > DCE_VERSION_11_0)
+	if (plane_state->ctx->dce_version > DCE_VERSION_MAX)
 		pipe_ctx->plane_res.scl_data.lb_params.depth = LB_PIXEL_DEPTH_36BPP;
 	else
 		pipe_ctx->plane_res.scl_data.lb_params.depth = LB_PIXEL_DEPTH_30BPP;

drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c

@@ -1228,6 +1228,8 @@ int smu_v11_0_set_fan_speed_rpm(struct smu_context *smu,
 	uint32_t crystal_clock_freq = 2500;
 	uint32_t tach_period;
 
+	if (speed == 0)
+		return -EINVAL;
 	/*
 	 * To prevent from possible overheat, some ASICs may have requirement
 	 * for minimum fan speed:

drivers/gpu/drm/i915/gem/i915_gem_ttm.c

@@ -620,10 +620,15 @@ i915_ttm_resource_get_st(struct drm_i915_gem_object *obj,
 			 struct ttm_resource *res)
 {
 	struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+	u64 page_alignment;
 
 	if (!i915_ttm_gtt_binds_lmem(res))
 		return i915_ttm_tt_get_st(bo->ttm);
 
+	page_alignment = bo->page_alignment << PAGE_SHIFT;
+	if (!page_alignment)
+		page_alignment = obj->mm.region->min_page_size;
+
 	/*
 	 * If CPU mapping differs, we need to add the ttm_tt pages to
 	 * the resulting st. Might make sense for GGTT.
@@ -634,7 +639,8 @@ i915_ttm_resource_get_st(struct drm_i915_gem_object *obj,
 			struct i915_refct_sgt *rsgt;
 
 			rsgt = intel_region_ttm_resource_to_rsgt(obj->mm.region,
-								 res);
+								 res,
+								 page_alignment);
 			if (IS_ERR(rsgt))
 				return rsgt;
 
@@ -643,7 +649,8 @@ i915_ttm_resource_get_st(struct drm_i915_gem_object *obj,
 		return i915_refct_sgt_get(obj->ttm.cached_io_rsgt);
 	}
 
-	return intel_region_ttm_resource_to_rsgt(obj->mm.region, res);
+	return intel_region_ttm_resource_to_rsgt(obj->mm.region, res,
+						 page_alignment);
 }
 
 static int i915_ttm_truncate(struct drm_i915_gem_object *obj)

drivers/gpu/drm/i915/gem/i915_gem_wait.c

@@ -9,6 +9,7 @@
 #include <linux/jiffies.h>
 
 #include "gt/intel_engine.h"
+#include "gt/intel_rps.h"
 
 #include "i915_gem_ioctls.h"
 #include "i915_gem_object.h"
@@ -31,6 +32,37 @@ i915_gem_object_wait_fence(struct dma_fence *fence,
 				      timeout);
 }
 
+static void
+i915_gem_object_boost(struct dma_resv *resv, unsigned int flags)
+{
+	struct dma_resv_iter cursor;
+	struct dma_fence *fence;
+
+	/*
+	 * Prescan all fences for potential boosting before we begin waiting.
+	 *
+	 * When we wait, we wait on outstanding fences serially. If the
+	 * dma-resv contains a sequence such as 1:1, 1:2 instead of a reduced
+	 * form 1:2, then as we look at each wait in turn we see that each
+	 * request is currently executing and not worthy of boosting. But if
+	 * we only happen to look at the final fence in the sequence (because
+	 * of request coalescing or splitting between read/write arrays by
+	 * the iterator), then we would boost. As such our decision to boost
+	 * or not is delicately balanced on the order we wait on fences.
+	 *
+	 * So instead of looking for boosts sequentially, look for all boosts
+	 * upfront and then wait on the outstanding fences.
+	 */
+
+	dma_resv_iter_begin(&cursor, resv,
+			    dma_resv_usage_rw(flags & I915_WAIT_ALL));
+	dma_resv_for_each_fence_unlocked(&cursor, fence)
+		if (dma_fence_is_i915(fence) &&
+		    !i915_request_started(to_request(fence)))
+			intel_rps_boost(to_request(fence));
+	dma_resv_iter_end(&cursor);
+}
+
 static long
 i915_gem_object_wait_reservation(struct dma_resv *resv,
 				 unsigned int flags,
@@ -40,6 +72,8 @@ i915_gem_object_wait_reservation(struct dma_resv *resv,
 	struct dma_fence *fence;
 	long ret = timeout ?: 1;
 
+	i915_gem_object_boost(resv, flags);
+
 	dma_resv_iter_begin(&cursor, resv,
 			    dma_resv_usage_rw(flags & I915_WAIT_ALL));
 	dma_resv_for_each_fence_unlocked(&cursor, fence) {

drivers/gpu/drm/i915/gt/intel_gt.c

@@ -1209,6 +1209,20 @@ void intel_gt_invalidate_tlbs(struct intel_gt *gt)
 	mutex_lock(&gt->tlb_invalidate_lock);
 	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
 
+	spin_lock_irq(&uncore->lock); /* serialise invalidate with GT reset */
+
+	for_each_engine(engine, gt, id) {
+		struct reg_and_bit rb;
+
+		rb = get_reg_and_bit(engine, regs == gen8_regs, regs, num);
+		if (!i915_mmio_reg_offset(rb.reg))
+			continue;
+
+		intel_uncore_write_fw(uncore, rb.reg, rb.bit);
+	}
+
+	spin_unlock_irq(&uncore->lock);
+
 	for_each_engine(engine, gt, id) {
 		/*
 		 * HW architecture suggest typical invalidation time at 40us,
@@ -1223,7 +1237,6 @@ void intel_gt_invalidate_tlbs(struct intel_gt *gt)
 		if (!i915_mmio_reg_offset(rb.reg))
 			continue;
 
-		intel_uncore_write_fw(uncore, rb.reg, rb.bit);
 		if (__intel_wait_for_register_fw(uncore,
 						 rb.reg, rb.bit, 0,
 						 timeout_us, timeout_ms,

drivers/gpu/drm/i915/gt/intel_reset.c

@@ -300,7 +300,7 @@ static int gen6_hw_domain_reset(struct intel_gt *gt, u32 hw_domain_mask)
 	return err;
 }
 
-static int gen6_reset_engines(struct intel_gt *gt,
+static int __gen6_reset_engines(struct intel_gt *gt,
 				intel_engine_mask_t engine_mask,
 				unsigned int retry)
 {
@@ -321,6 +321,20 @@ static int gen6_reset_engines(struct intel_gt *gt,
 	return gen6_hw_domain_reset(gt, hw_mask);
 }
 
+static int gen6_reset_engines(struct intel_gt *gt,
+			      intel_engine_mask_t engine_mask,
+			      unsigned int retry)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&gt->uncore->lock, flags);
+	ret = __gen6_reset_engines(gt, engine_mask, retry);
+	spin_unlock_irqrestore(&gt->uncore->lock, flags);
+
+	return ret;
+}
+
 static struct intel_engine_cs *find_sfc_paired_vecs_engine(struct intel_engine_cs *engine)
 {
 	int vecs_id;
@@ -487,7 +501,7 @@ static void gen11_unlock_sfc(struct intel_engine_cs *engine)
 	rmw_clear_fw(uncore, sfc_lock.lock_reg, sfc_lock.lock_bit);
 }
 
-static int gen11_reset_engines(struct intel_gt *gt,
+static int __gen11_reset_engines(struct intel_gt *gt,
 				 intel_engine_mask_t engine_mask,
 				 unsigned int retry)
 {
@@ -583,8 +597,11 @@ static int gen8_reset_engines(struct intel_gt *gt,
 	struct intel_engine_cs *engine;
 	const bool reset_non_ready = retry >= 1;
 	intel_engine_mask_t tmp;
+	unsigned long flags;
 	int ret;
 
+	spin_lock_irqsave(&gt->uncore->lock, flags);
+
 	for_each_engine_masked(engine, gt, engine_mask, tmp) {
 		ret = gen8_engine_reset_prepare(engine);
 		if (ret && !reset_non_ready)
@@ -612,17 +629,19 @@ static int gen8_reset_engines(struct intel_gt *gt,
 	 * This is best effort, so ignore any error from the initial reset.
 	 */
 	if (IS_DG2(gt->i915) && engine_mask == ALL_ENGINES)
-		gen11_reset_engines(gt, gt->info.engine_mask, 0);
+		__gen11_reset_engines(gt, gt->info.engine_mask, 0);
 
 	if (GRAPHICS_VER(gt->i915) >= 11)
-		ret = gen11_reset_engines(gt, engine_mask, retry);
+		ret = __gen11_reset_engines(gt, engine_mask, retry);
 	else
-		ret = gen6_reset_engines(gt, engine_mask, retry);
+		ret = __gen6_reset_engines(gt, engine_mask, retry);
 
 skip_reset:
 	for_each_engine_masked(engine, gt, engine_mask, tmp)
 		gen8_engine_reset_cancel(engine);
 
+	spin_unlock_irqrestore(&gt->uncore->lock, flags);
+
 	return ret;
 }
 

drivers/gpu/drm/i915/gt/selftest_lrc.c

@@ -176,8 +176,8 @@ static int live_lrc_layout(void *arg)
 			continue;
 
 		hw = shmem_pin_map(engine->default_state);
-		if (IS_ERR(hw)) {
-			err = PTR_ERR(hw);
+		if (!hw) {
+			err = -ENOMEM;
 			break;
 		}
 		hw += LRC_STATE_OFFSET / sizeof(*hw);
@@ -365,8 +365,8 @@ static int live_lrc_fixed(void *arg)
 			continue;
 
 		hw = shmem_pin_map(engine->default_state);
-		if (IS_ERR(hw)) {
-			err = PTR_ERR(hw);
+		if (!hw) {
+			err = -ENOMEM;
 			break;
 		}
 		hw += LRC_STATE_OFFSET / sizeof(*hw);

drivers/gpu/drm/i915/gvt/cmd_parser.c

@@ -3117,9 +3117,9 @@ void intel_gvt_update_reg_whitelist(struct intel_vgpu *vgpu)
 			continue;
 
 		vaddr = shmem_pin_map(engine->default_state);
-		if (IS_ERR(vaddr)) {
-			gvt_err("failed to map %s->default state, err:%zd\n",
-				engine->name, PTR_ERR(vaddr));
+		if (!vaddr) {
+			gvt_err("failed to map %s->default state\n",
+				engine->name);
 			return;
 		}
 

drivers/gpu/drm/i915/i915_scatterlist.c

@@ -68,6 +68,7 @@ void i915_refct_sgt_init(struct i915_refct_sgt *rsgt, size_t size)
  * drm_mm_node
  * @node: The drm_mm_node.
  * @region_start: An offset to add to the dma addresses of the sg list.
+ * @page_alignment: Required page alignment for each sg entry. Power of two.
  *
  * Create a struct sg_table, initializing it from a struct drm_mm_node,
  * taking a maximum segment length into account, splitting into segments
@@ -77,15 +78,18 @@ void i915_refct_sgt_init(struct i915_refct_sgt *rsgt, size_t size)
  * error code cast to an error pointer on failure.
  */
 struct i915_refct_sgt *i915_rsgt_from_mm_node(const struct drm_mm_node *node,
-					      u64 region_start)
+					      u64 region_start,
+					      u64 page_alignment)
 {
-	const u64 max_segment = SZ_1G; /* Do we have a limit on this? */
+	const u64 max_segment = round_down(UINT_MAX, page_alignment);
 	u64 segment_pages = max_segment >> PAGE_SHIFT;
 	u64 block_size, offset, prev_end;
 	struct i915_refct_sgt *rsgt;
 	struct sg_table *st;
 	struct scatterlist *sg;
 
+	GEM_BUG_ON(!max_segment);
+
 	rsgt = kmalloc(sizeof(*rsgt), GFP_KERNEL);
 	if (!rsgt)
 		return ERR_PTR(-ENOMEM);
@@ -112,6 +116,8 @@ struct i915_refct_sgt *i915_rsgt_from_mm_node(const struct drm_mm_node *node,
 				sg = __sg_next(sg);
 
 			sg_dma_address(sg) = region_start + offset;
+			GEM_BUG_ON(!IS_ALIGNED(sg_dma_address(sg),
+					       page_alignment));
 			sg_dma_len(sg) = 0;
 			sg->length = 0;
 			st->nents++;
@@ -138,6 +144,7 @@ struct i915_refct_sgt *i915_rsgt_from_mm_node(const struct drm_mm_node *node,
  * i915_buddy_block list
  * @res: The struct i915_ttm_buddy_resource.
  * @region_start: An offset to add to the dma addresses of the sg list.
+ * @page_alignment: Required page alignment for each sg entry. Power of two.
  *
  * Create a struct sg_table, initializing it from struct i915_buddy_block list,
  * taking a maximum segment length into account, splitting into segments
@@ -147,11 +154,12 @@ struct i915_refct_sgt *i915_rsgt_from_mm_node(const struct drm_mm_node *node,
  * error code cast to an error pointer on failure.
  */
 struct i915_refct_sgt *i915_rsgt_from_buddy_resource(struct ttm_resource *res,
-						     u64 region_start)
+						     u64 region_start,
+						     u64 page_alignment)
 {
 	struct i915_ttm_buddy_resource *bman_res = to_ttm_buddy_resource(res);
 	const u64 size = res->num_pages << PAGE_SHIFT;
-	const u64 max_segment = rounddown(UINT_MAX, PAGE_SIZE);
+	const u64 max_segment = round_down(UINT_MAX, page_alignment);
 	struct drm_buddy *mm = bman_res->mm;
 	struct list_head *blocks = &bman_res->blocks;
 	struct drm_buddy_block *block;
@@ -161,6 +169,7 @@ struct i915_refct_sgt *i915_rsgt_from_buddy_resource(struct ttm_resource *res,
 	resource_size_t prev_end;
 
 	GEM_BUG_ON(list_empty(blocks));
+	GEM_BUG_ON(!max_segment);
 
 	rsgt = kmalloc(sizeof(*rsgt), GFP_KERNEL);
 	if (!rsgt)
@@ -191,6 +200,8 @@ struct i915_refct_sgt *i915_rsgt_from_buddy_resource(struct ttm_resource *res,
 					sg = __sg_next(sg);
 
 				sg_dma_address(sg) = region_start + offset;
+				GEM_BUG_ON(!IS_ALIGNED(sg_dma_address(sg),
+						       page_alignment));
 				sg_dma_len(sg) = 0;
 				sg->length = 0;
 				st->nents++;

drivers/gpu/drm/i915/i915_scatterlist.h

@@ -213,9 +213,11 @@ static inline void __i915_refct_sgt_init(struct i915_refct_sgt *rsgt,
 void i915_refct_sgt_init(struct i915_refct_sgt *rsgt, size_t size);
 
 struct i915_refct_sgt *i915_rsgt_from_mm_node(const struct drm_mm_node *node,
-					      u64 region_start);
+					      u64 region_start,
+					      u64 page_alignment);
 
 struct i915_refct_sgt *i915_rsgt_from_buddy_resource(struct ttm_resource *res,
-						     u64 region_start);
+						     u64 region_start,
+						     u64 page_alignment);
 
 #endif

drivers/gpu/drm/i915/intel_region_ttm.c

@@ -152,6 +152,7 @@ int intel_region_ttm_fini(struct intel_memory_region *mem)
  * Convert an opaque TTM resource manager resource to a refcounted sg_table.
  * @mem: The memory region.
  * @res: The resource manager resource obtained from the TTM resource manager.
+ * @page_alignment: Required page alignment for each sg entry. Power of two.
  *
  * The gem backends typically use sg-tables for operations on the underlying
  * io_memory. So provide a way for the backends to translate the
@@ -161,16 +162,19 @@ int intel_region_ttm_fini(struct intel_memory_region *mem)
  */
 struct i915_refct_sgt *
 intel_region_ttm_resource_to_rsgt(struct intel_memory_region *mem,
-				  struct ttm_resource *res)
+				  struct ttm_resource *res,
+				  u64 page_alignment)
 {
 	if (mem->is_range_manager) {
 		struct ttm_range_mgr_node *range_node =
 			to_ttm_range_mgr_node(res);
 
 		return i915_rsgt_from_mm_node(&range_node->mm_nodes[0],
-					      mem->region.start);
+					      mem->region.start,
+					      page_alignment);
 	} else {
-		return i915_rsgt_from_buddy_resource(res, mem->region.start);
+		return i915_rsgt_from_buddy_resource(res, mem->region.start,
+						     page_alignment);
 	}
 }
 

drivers/gpu/drm/i915/intel_region_ttm.h

@@ -24,7 +24,8 @@ int intel_region_ttm_fini(struct intel_memory_region *mem);
 
 struct i915_refct_sgt *
 intel_region_ttm_resource_to_rsgt(struct intel_memory_region *mem,
-				  struct ttm_resource *res);
+				  struct ttm_resource *res,
+				  u64 page_alignment);
 
 void intel_region_ttm_resource_free(struct intel_memory_region *mem,
 				    struct ttm_resource *res);

drivers/gpu/drm/i915/selftests/i915_gem_gtt.c

@@ -742,7 +742,7 @@ static int pot_hole(struct i915_address_space *vm,
 		u64 addr;
 
 		for (addr = round_up(hole_start + min_alignment, step) - min_alignment;
-		     addr <= round_down(hole_end - (2 * min_alignment), step) - min_alignment;
+		     hole_end > addr && hole_end - addr >= 2 * min_alignment;
 		     addr += step) {
 			err = i915_vma_pin(vma, 0, 0, addr | flags);
 			if (err) {

drivers/gpu/drm/i915/selftests/intel_memory_region.c

@@ -451,7 +451,6 @@ static int igt_mock_splintered_region(void *arg)
 
 static int igt_mock_max_segment(void *arg)
 {
-	const unsigned int max_segment = rounddown(UINT_MAX, PAGE_SIZE);
 	struct intel_memory_region *mem = arg;
 	struct drm_i915_private *i915 = mem->i915;
 	struct i915_ttm_buddy_resource *res;
@@ -460,7 +459,10 @@ static int igt_mock_max_segment(void *arg)
 	struct drm_buddy *mm;
 	struct list_head *blocks;
 	struct scatterlist *sg;
+	I915_RND_STATE(prng);
 	LIST_HEAD(objects);
+	unsigned int max_segment;
+	unsigned int ps;
 	u64 size;
 	int err = 0;
 
@@ -472,7 +474,13 @@ static int igt_mock_max_segment(void *arg)
 	 */
 
 	size = SZ_8G;
-	mem = mock_region_create(i915, 0, size, PAGE_SIZE, 0, 0);
+	ps = PAGE_SIZE;
+	if (i915_prandom_u64_state(&prng) & 1)
+		ps = SZ_64K; /* For something like DG2 */
+
+	max_segment = round_down(UINT_MAX, ps);
+
+	mem = mock_region_create(i915, 0, size, ps, 0, 0);
 	if (IS_ERR(mem))
 		return PTR_ERR(mem);
 
@@ -498,12 +506,21 @@ static int igt_mock_max_segment(void *arg)
 	}
 
 	for (sg = obj->mm.pages->sgl; sg; sg = sg_next(sg)) {
+		dma_addr_t daddr = sg_dma_address(sg);
+
 		if (sg->length > max_segment) {
 			pr_err("%s: Created an oversized scatterlist entry, %u > %u\n",
 			       __func__, sg->length, max_segment);
 			err = -EINVAL;
 			goto out_close;
 		}
+
+		if (!IS_ALIGNED(daddr, ps)) {
+			pr_err("%s: Created an unaligned scatterlist entry, addr=%pa, ps=%u\n",
+			       __func__,  &daddr, ps);
+			err = -EINVAL;
+			goto out_close;
+		}
 	}
 
 out_close:

drivers/gpu/drm/i915/selftests/mock_region.c

@@ -33,7 +33,8 @@ static int mock_region_get_pages(struct drm_i915_gem_object *obj)
 		return PTR_ERR(obj->mm.res);
 
 	obj->mm.rsgt = intel_region_ttm_resource_to_rsgt(obj->mm.region,
-							 obj->mm.res);
+							 obj->mm.res,
+							 obj->mm.region->min_page_size);
 	if (IS_ERR(obj->mm.rsgt)) {
 		err = PTR_ERR(obj->mm.rsgt);
 		goto err_free_resource;