Merge tag 'drm-xe-next-2024-12-11' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next

UAPI Changes:
 - Make OA buffer size configurable (Sai)

Display Changes (including i915):
 - Fix ttm_bo_access() usage (Auld)
 - Power request asserting/deasserting for Xe3lpd (Mika)
 - One Type-C conversion towards struct intel_display (Mika)

Driver Changes:
 - GuC capture related fixes (Everest, Zhanjun)
 - Move old workaround to OOB infra (Lucas)
 - Compute mode change refactoring (Bala)
 - Add ufence and g2h flushes for LNL Hybrid timeouts (Nirmoy)
 - Avoid unnecessary OOM kills (Thomas)
 - Restore system memory GGTT mappings (Brost)
 - Fix build error for XE_IOCTL_DBG macro (Gyeyoung)
 - Documentation updates and fixes (Lucas, Randy)
 - A few exec IOCTL fixes (Brost)
 - Fix potential GGTT allocation leak (Michal)
 - Fix races on fdinfo (Lucas)
 - SRIOV VF: Post-migration recovery worker basis (Tomasz)
 - GuC Communication fixes and improvements (Michal, John, Tomasz, Auld, Jonathan)
 - SRIOV PF: Add support for VF scheduling priority
 - Trace improvements (Lucas, Auld, Oak)
 - Hibernation on igpu fixes and improvements (Auld)
 - GT oriented logs/asserts improvements (Michal)
 - Take job list lock in xe_sched_first_pending_job (Nirmoy)
 - GSC: Improve SW proxy error checking and logging (Daniele)
 - GuC crash notifications & drop default log verbosity (John)
 - Fix races on fdinfo (Lucas)
 - Fix runtime_pm handling in OA (Ashutosh)
 - Allow fault injection in vm create and vm bind IOCTLs (Francois)
 - TLB invalidation fixes (Nirmoy, Daniele)
 - Devcoredump Improvements, doc and fixes (Brost, Lucas, Zhanjun, John)
 - Wake up waiters after setting ufence->signalled (Nirmoy)
 - Mark preempt fence workqueue as reclaim (Brost)
 - Trivial header/flags cleanups (Lucas)
 - VRAM drop 2G block restriction (Auld)
 - Drop useless d3cold allowed message (Brost)
 - SRIOV PF: Drop 2GiB limit of fair LMEM allocation (Michal)
 - Add another PTL PCI ID (Atwood)
 - Allow bo mapping on multiple ggtts (Niranjana)
 - Add support for GuC-to-GuC communication (John)
 - Update xe2_graphics name string (Roper)
 - VRAM: fix lpfn check (Auld)
 - Ad Xe3 workaround (Apoorva)
 - Migrate fixes (Auld)
 - Fix non-contiguous VRAM BO access (Brost)
 - Log throttle reasons (Raag)
 - Enable PMT support for BMG (Michael)
 - IRQ related fixes and improvements (Ilia)
 - Avoid evicting object of the same vm in none fault mode (Oak)
 - Fix in tests (Nirmoy)
 - Fix ERR_PTR handling (Mirsad)
 - Some reg_sr/whitelist fixes and refactors (Lucas)

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

# -----BEGIN PGP SIGNATURE-----
#
# iQEzBAABCAAdFiEEbSBwaO7dZQkcLOKj+mJfZA7rE8oFAmdaHkMACgkQ+mJfZA7r
# E8o+twf/XYZTk4O3qQ+yNL3PDQT0NIKjH8mEnmu4udyIw/sYhQe6ji+uh1YutK8Y
# 41IQc06qQogTj36bqSwbjThw5asMfRh2sNR/p1uOy7RGUnN25FuYSXEgOeDWi/Ec
# xrZE1TKPotFGeGI09KJmzjzMq94cgv97Pxma+5m8BjVsvzXQSzEJ2r9cC6ruSfNT
# O5Jq5nqxHSkWUbKCxPnixSlGnH4jbsuiqS1E1pnH+u6ijxsfhOJj686wLn2FRkiw
# 6FhXmJBrd8AZ0Q2E7h3UswE5O88I0ALDc58OINAzD1GMyzvZj2vB1pXgj5uNr0/x
# Ku4cxu1jprsi+FLUdKAdYpxRBRanow==
# =3Ou7
# -----END PGP SIGNATURE-----
# gpg: Signature made Thu 12 Dec 2024 09:20:35 AEST
# gpg:                using RSA key 6D207068EEDD65091C2CE2A3FA625F640EEB13CA
# gpg: Good signature from "Rodrigo Vivi <rodrigo.vivi@intel.com>" [unknown]
# gpg:                 aka "Rodrigo Vivi <rodrigo.vivi@gmail.com>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 6D20 7068 EEDD 6509 1C2C  E2A3 FA62 5F64 0EEB 13CA
From: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/Z1ofx-fExLQKV_e4@intel.com

94 files changed, 2461 insertions, 588 deletions

diff --git a/Documentation/gpu/xe/index.rst b/Documentation/gpu/xe/index.rst
index 3f07aa3b5432..92cfb25e64d3 100644
--- a/Documentation/gpu/xe/index.rst
+++ b/Documentation/gpu/xe/index.rst
@@ -23,4 +23,5 @@ DG2, etc is provided to prototype the driver.
    xe_firmware
    xe_tile
    xe_debugging
+   xe_devcoredump
    xe-drm-usage-stats.rst
diff --git a/Documentation/gpu/xe/xe_devcoredump.rst b/Documentation/gpu/xe/xe_devcoredump.rst
new file mode 100644
index 000000000000..ae4ec0e34dc0
--- /dev/null
+++ b/Documentation/gpu/xe/xe_devcoredump.rst
@@ -0,0 +1,14 @@
+.. SPDX-License-Identifier: (GPL-2.0+ OR MIT)
+
+==================
+Xe Device Coredump
+==================
+
+.. kernel-doc:: drivers/gpu/drm/xe/xe_devcoredump.c
+   :doc: Xe device coredump
+
+Internal API
+============
+
+.. kernel-doc:: drivers/gpu/drm/xe/xe_devcoredump.c
+   :internal:
diff --git a/drivers/gpu/drm/i915/display/intel_cx0_phy_regs.h b/drivers/gpu/drm/i915/display/intel_cx0_phy_regs.h
index f0e5c196eae4..c685479c9756 100644
--- a/drivers/gpu/drm/i915/display/intel_cx0_phy_regs.h
+++ b/drivers/gpu/drm/i915/display/intel_cx0_phy_regs.h
@@ -200,6 +200,13 @@
 #define   XELPDP_SSC_ENABLE_PLLA			REG_BIT(1)
 #define   XELPDP_SSC_ENABLE_PLLB			REG_BIT(0)
 
+#define TCSS_DISP_MAILBOX_IN_CMD		_MMIO(0x161300)
+#define   TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY	REG_BIT(31)
+#define   TCSS_DISP_MAILBOX_IN_CMD_CMD_MASK	REG_GENMASK(7, 0)
+#define   TCSS_DISP_MAILBOX_IN_CMD_DATA(val)	REG_FIELD_PREP(TCSS_DISP_MAILBOX_IN_CMD_CMD_MASK, val)
+
+#define TCSS_DISP_MAILBOX_IN_DATA		_MMIO(0x161304)
+
 /* C10 Vendor Registers */
 #define PHY_C10_VDR_PLL(idx)		(0xC00 + (idx))
 #define   C10_PLL0_FRACEN		REG_BIT8(4)
diff --git a/drivers/gpu/drm/i915/display/intel_tc.c b/drivers/gpu/drm/i915/display/intel_tc.c
index b16c4d2d4077..33b8f2772076 100644
--- a/drivers/gpu/drm/i915/display/intel_tc.c
+++ b/drivers/gpu/drm/i915/display/intel_tc.c
@@ -1013,21 +1013,52 @@ xelpdp_tc_phy_wait_for_tcss_power(struct intel_tc_port *tc, bool enabled)
 	return true;
 }
 
+/*
+ * Gfx driver WA 14020908590 for PTL tcss_rxdetect_clkswb_req/ack
+ * handshake violation when pwwreq= 0->1 during TC7/10 entry
+ */
+static void xelpdp_tc_power_request_wa(struct intel_display *display, bool enable)
+{
+	/* check if mailbox is running busy */
+	if (intel_de_wait_for_clear(display, TCSS_DISP_MAILBOX_IN_CMD,
+				    TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY, 10)) {
+		drm_dbg_kms(display->drm,
+			    "Timeout waiting for TCSS mailbox run/busy bit to clear\n");
+		return;
+	}
+
+	intel_de_write(display, TCSS_DISP_MAILBOX_IN_DATA, enable ? 1 : 0);
+	intel_de_write(display, TCSS_DISP_MAILBOX_IN_CMD,
+		       TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY |
+		       TCSS_DISP_MAILBOX_IN_CMD_DATA(0x1));
+
+	/* wait to clear mailbox running busy bit before continuing */
+	if (intel_de_wait_for_clear(display, TCSS_DISP_MAILBOX_IN_CMD,
+				    TCSS_DISP_MAILBOX_IN_CMD_RUN_BUSY, 10)) {
+		drm_dbg_kms(display->drm,
+			    "Timeout after writing data to mailbox. Mailbox run/busy bit did not clear\n");
+		return;
+	}
+}
+
 static void __xelpdp_tc_phy_enable_tcss_power(struct intel_tc_port *tc, bool enable)
 {
-	struct drm_i915_private *i915 = tc_to_i915(tc);
+	struct intel_display *display = to_intel_display(tc->dig_port);
 	enum port port = tc->dig_port->base.port;
-	i915_reg_t reg = XELPDP_PORT_BUF_CTL1(i915, port);
+	i915_reg_t reg = XELPDP_PORT_BUF_CTL1(display, port);
 	u32 val;
 
 	assert_tc_cold_blocked(tc);
 
-	val = intel_de_read(i915, reg);
+	if (DISPLAY_VER(display) == 30)
+		xelpdp_tc_power_request_wa(display, enable);
+
+	val = intel_de_read(display, reg);
 	if (enable)
 		val |= XELPDP_TCSS_POWER_REQUEST;
 	else
 		val &= ~XELPDP_TCSS_POWER_REQUEST;
-	intel_de_write(i915, reg, val);
+	intel_de_write(display, reg, val);
 }
 
 static bool xelpdp_tc_phy_enable_tcss_power(struct intel_tc_port *tc, bool enable)
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 0cf91d23f25c..a194db83421d 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -405,13 +405,25 @@ static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo,
 	return len;
 }
 
-int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
-		     void *buf, int len, int write)
+/**
+ * ttm_bo_access - Helper to access a buffer object
+ *
+ * @bo: ttm buffer object
+ * @offset: access offset into buffer object
+ * @buf: pointer to caller memory to read into or write from
+ * @len: length of access
+ * @write: write access
+ *
+ * Utility function to access a buffer object. Useful when buffer object cannot
+ * be easily mapped (non-contiguous, non-visible, etc...). Should not directly
+ * be exported to user space via a peak / poke interface.
+ *
+ * Returns:
+ * @len if successful, negative error code on failure.
+ */
+int ttm_bo_access(struct ttm_buffer_object *bo, unsigned long offset,
+		  void *buf, int len, int write)
 {
-	struct ttm_buffer_object *bo = vma->vm_private_data;
-	unsigned long offset = (addr) - vma->vm_start +
-		((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
-		 << PAGE_SHIFT);
 	int ret;
 
 	if (len < 1 || (offset + len) > bo->base.size)
@@ -429,8 +441,8 @@ int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
 		break;
 	default:
 		if (bo->bdev->funcs->access_memory)
-			ret = bo->bdev->funcs->access_memory(
-				bo, offset, buf, len, write);
+			ret = bo->bdev->funcs->access_memory
+				(bo, offset, buf, len, write);
 		else
 			ret = -EIO;
 	}
@@ -439,6 +451,18 @@ int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
 
 	return ret;
 }
+EXPORT_SYMBOL(ttm_bo_access);
+
+int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
+		     void *buf, int len, int write)
+{
+	struct ttm_buffer_object *bo = vma->vm_private_data;
+	unsigned long offset = (addr) - vma->vm_start +
+		((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+		 << PAGE_SHIFT);
+
+	return ttm_bo_access(bo, offset, buf, len, write);
+}
 EXPORT_SYMBOL(ttm_bo_vm_access);
 
 static const struct vm_operations_struct ttm_bo_vm_ops = {
diff --git a/drivers/gpu/drm/xe/Makefile b/drivers/gpu/drm/xe/Makefile
index bc7a04ce69fd..7730e0596299 100644
--- a/drivers/gpu/drm/xe/Makefile
+++ b/drivers/gpu/drm/xe/Makefile
@@ -101,6 +101,7 @@ xe-y += xe_bb.o \
 	xe_trace.o \
 	xe_trace_bo.o \
 	xe_trace_guc.o \
+	xe_trace_lrc.o \
 	xe_ttm_sys_mgr.o \
 	xe_ttm_stolen_mgr.o \
 	xe_ttm_vram_mgr.o \
@@ -110,6 +111,7 @@ xe-y += xe_bb.o \
 	xe_vm.o \
 	xe_vram.o \
 	xe_vram_freq.o \
+	xe_vsec.o \
 	xe_wait_user_fence.o \
 	xe_wa.o \
 	xe_wopcm.o
@@ -124,7 +126,8 @@ xe-y += \
 	xe_gt_sriov_vf.o \
 	xe_guc_relay.o \
 	xe_memirq.o \
-	xe_sriov.o
+	xe_sriov.o \
+	xe_sriov_vf.o
 
 xe-$(CONFIG_PCI_IOV) += \
 	xe_gt_sriov_pf.o \
diff --git a/drivers/gpu/drm/xe/abi/guc_actions_abi.h b/drivers/gpu/drm/xe/abi/guc_actions_abi.h
index b54fe40fc5a9..fee385532fb0 100644
--- a/drivers/gpu/drm/xe/abi/guc_actions_abi.h
+++ b/drivers/gpu/drm/xe/abi/guc_actions_abi.h
@@ -134,6 +134,8 @@ enum xe_guc_action {
 	XE_GUC_ACTION_DEREGISTER_CONTEXT = 0x4503,
 	XE_GUC_ACTION_REGISTER_COMMAND_TRANSPORT_BUFFER = 0x4505,
 	XE_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER = 0x4506,
+	XE_GUC_ACTION_REGISTER_G2G = 0x4507,
+	XE_GUC_ACTION_DEREGISTER_G2G = 0x4508,
 	XE_GUC_ACTION_DEREGISTER_CONTEXT_DONE = 0x4600,
 	XE_GUC_ACTION_REGISTER_CONTEXT_MULTI_LRC = 0x4601,
 	XE_GUC_ACTION_CLIENT_SOFT_RESET = 0x5507,
@@ -218,4 +220,22 @@ enum xe_guc_tlb_inval_mode {
 	XE_GUC_TLB_INVAL_MODE_LITE = 0x1,
 };
 
+/*
+ * GuC to GuC communication (de-)registration fields:
+ */
+enum xe_guc_g2g_type {
+	XE_G2G_TYPE_IN = 0x0,
+	XE_G2G_TYPE_OUT,
+	XE_G2G_TYPE_LIMIT,
+};
+
+#define XE_G2G_REGISTER_DEVICE	REG_GENMASK(16, 16)
+#define XE_G2G_REGISTER_TILE	REG_GENMASK(15, 12)
+#define XE_G2G_REGISTER_TYPE	REG_GENMASK(11, 8)
+#define XE_G2G_REGISTER_SIZE	REG_GENMASK(7, 0)
+
+#define XE_G2G_DEREGISTER_DEVICE	REG_GENMASK(16, 16)
+#define XE_G2G_DEREGISTER_TILE	REG_GENMASK(15, 12)
+#define XE_G2G_DEREGISTER_TYPE	REG_GENMASK(11, 8)
+
 #endif
diff --git a/drivers/gpu/drm/xe/abi/guc_actions_sriov_abi.h b/drivers/gpu/drm/xe/abi/guc_actions_sriov_abi.h
index b6a1852749dd..0b28659d94e9 100644
--- a/drivers/gpu/drm/xe/abi/guc_actions_sriov_abi.h
+++ b/drivers/gpu/drm/xe/abi/guc_actions_sriov_abi.h
@@ -502,6 +502,44 @@
 #define VF2GUC_VF_RESET_RESPONSE_MSG_0_MBZ		GUC_HXG_RESPONSE_MSG_0_DATA0
 
 /**
+ * DOC: VF2GUC_NOTIFY_RESFIX_DONE
+ *
+ * This action is used by VF to notify the GuC that the VF KMD has completed
+ * post-migration recovery steps.
+ *
+ * This message must be sent as `MMIO HXG Message`_.
+ *
+ *  +---+-------+--------------------------------------------------------------+
+ *  |   | Bits  | Description                                                  |
+ *  +===+=======+==============================================================+
+ *  | 0 |    31 | ORIGIN = GUC_HXG_ORIGIN_HOST_                                |
+ *  |   +-------+--------------------------------------------------------------+
+ *  |   | 30:28 | TYPE = GUC_HXG_TYPE_REQUEST_                                 |
+ *  |   +-------+--------------------------------------------------------------+
+ *  |   | 27:16 | DATA0 = MBZ                                                  |
+ *  |   +-------+--------------------------------------------------------------+
+ *  |   |  15:0 | ACTION = _`GUC_ACTION_VF2GUC_NOTIFY_RESFIX_DONE` = 0x5508    |
+ *  +---+-------+--------------------------------------------------------------+
+ *
+ *  +---+-------+--------------------------------------------------------------+
+ *  |   | Bits  | Description                                                  |
+ *  +===+=======+==============================================================+
+ *  | 0 |    31 | ORIGIN = GUC_HXG_ORIGIN_GUC_                                 |
+ *  |   +-------+--------------------------------------------------------------+
+ *  |   | 30:28 | TYPE = GUC_HXG_TYPE_RESPONSE_SUCCESS_                        |
+ *  |   +-------+--------------------------------------------------------------+
+ *  |   |  27:0 | DATA0 = MBZ                                                  |
+ *  +---+-------+--------------------------------------------------------------+
+ */
+#define GUC_ACTION_VF2GUC_NOTIFY_RESFIX_DONE		0x5508u
+
+#define VF2GUC_NOTIFY_RESFIX_DONE_REQUEST_MSG_LEN	GUC_HXG_REQUEST_MSG_MIN_LEN
+#define VF2GUC_NOTIFY_RESFIX_DONE_REQUEST_MSG_0_MBZ	GUC_HXG_REQUEST_MSG_0_DATA0
+
+#define VF2GUC_NOTIFY_RESFIX_DONE_RESPONSE_MSG_LEN	GUC_HXG_RESPONSE_MSG_MIN_LEN
+#define VF2GUC_NOTIFY_RESFIX_DONE_RESPONSE_MSG_0_MBZ	GUC_HXG_RESPONSE_MSG_0_DATA0
+
+/**
  * DOC: VF2GUC_QUERY_SINGLE_KLV
  *
  * This action is used by VF to query value of the single KLV data.
diff --git a/drivers/gpu/drm/xe/abi/guc_klvs_abi.h b/drivers/gpu/drm/xe/abi/guc_klvs_abi.h
index 37606cf8cc5e..7dcb118e3d9f 100644
--- a/drivers/gpu/drm/xe/abi/guc_klvs_abi.h
+++ b/drivers/gpu/drm/xe/abi/guc_klvs_abi.h
@@ -291,6 +291,14 @@ enum  {
  *
  *      :0: (default)
  *      :1-65535: number of contexts (Gen12)
+ *
+ * _`GUC_KLV_VF_CFG_SCHED_PRIORITY` : 0x8A0C
+ *      This config controls VF’s scheduling priority.
+ *
+ *      :0: LOW = schedule VF only if it has active work (default)
+ *      :1: NORMAL = schedule VF always, irrespective of whether it has work or not
+ *      :2: HIGH = schedule VF in the next time-slice after current active
+ *          time-slice completes if it has active work
  */
 
 #define GUC_KLV_VF_CFG_GGTT_START_KEY		0x0001
@@ -343,6 +351,12 @@ enum  {
 #define GUC_KLV_VF_CFG_BEGIN_CONTEXT_ID_KEY	0x8a0b
 #define GUC_KLV_VF_CFG_BEGIN_CONTEXT_ID_LEN	1u
 
+#define GUC_KLV_VF_CFG_SCHED_PRIORITY_KEY	0x8a0c
+#define GUC_KLV_VF_CFG_SCHED_PRIORITY_LEN	1u
+#define   GUC_SCHED_PRIORITY_LOW		0u
+#define   GUC_SCHED_PRIORITY_NORMAL		1u
+#define   GUC_SCHED_PRIORITY_HIGH		2u
+
 /*
  * Workaround keys:
  */
diff --git a/drivers/gpu/drm/xe/display/ext/i915_irq.c b/drivers/gpu/drm/xe/display/ext/i915_irq.c
index a7dbc6554d69..ac4cda2d81c7 100644
--- a/drivers/gpu/drm/xe/display/ext/i915_irq.c
+++ b/drivers/gpu/drm/xe/display/ext/i915_irq.c
@@ -53,18 +53,7 @@ void gen2_irq_init(struct intel_uncore *uncore, struct i915_irq_regs regs,
 
 bool intel_irqs_enabled(struct xe_device *xe)
 {
-	/*
-	 * XXX: i915 has a racy handling of the irq.enabled, since it doesn't
-	 * lock its transitions. Because of that, the irq.enabled sometimes
-	 * is not read with the irq.lock in place.
-	 * However, the most critical cases like vblank and page flips are
-	 * properly using the locks.
-	 * We cannot take the lock in here or run any kind of assert because
-	 * of i915 inconsistency.
-	 * But at this point the xe irq is better protected against races,
-	 * although the full solution would be protecting the i915 side.
-	 */
-	return xe->irq.enabled;
+	return atomic_read(&xe->irq.enabled);
 }
 
 void intel_synchronize_irq(struct xe_device *xe)
diff --git a/drivers/gpu/drm/xe/display/intel_bo.c b/drivers/gpu/drm/xe/display/intel_bo.c
index 9f54fad0f1c0..b463f5bd4eed 100644
--- a/drivers/gpu/drm/xe/display/intel_bo.c
+++ b/drivers/gpu/drm/xe/display/intel_bo.c
@@ -40,31 +40,8 @@ int intel_bo_fb_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
 int intel_bo_read_from_page(struct drm_gem_object *obj, u64 offset, void *dst, int size)
 {
 	struct xe_bo *bo = gem_to_xe_bo(obj);
-	struct ttm_bo_kmap_obj map;
-	void *src;
-	bool is_iomem;
-	int ret;
 
-	ret = xe_bo_lock(bo, true);
-	if (ret)
-		return ret;
-
-	ret = ttm_bo_kmap(&bo->ttm, offset >> PAGE_SHIFT, 1, &map);
-	if (ret)
-		goto out_unlock;
-
-	offset &= ~PAGE_MASK;
-	src = ttm_kmap_obj_virtual(&map, &is_iomem);
-	src += offset;
-	if (is_iomem)
-		memcpy_fromio(dst, (void __iomem *)src, size);
-	else
-		memcpy(dst, src, size);
-
-	ttm_bo_kunmap(&map);
-out_unlock:
-	xe_bo_unlock(bo);
-	return ret;
+	return xe_bo_read(bo, offset, dst, size);
 }
 
 struct intel_frontbuffer *intel_bo_get_frontbuffer(struct drm_gem_object *obj)
diff --git a/drivers/gpu/drm/xe/display/xe_fb_pin.c b/drivers/gpu/drm/xe/display/xe_fb_pin.c
index 761510ae0690..9fa51b84737c 100644
--- a/drivers/gpu/drm/xe/display/xe_fb_pin.c
+++ b/drivers/gpu/drm/xe/display/xe_fb_pin.c
@@ -161,7 +161,7 @@ static int __xe_pin_fb_vma_dpt(const struct intel_framebuffer *fb,
 	}
 
 	vma->dpt = dpt;
-	vma->node = dpt->ggtt_node;
+	vma->node = dpt->ggtt_node[tile0->id];
 	return 0;
 }
 
@@ -213,8 +213,8 @@ static int __xe_pin_fb_vma_ggtt(const struct intel_framebuffer *fb,
 	if (xe_bo_is_vram(bo) && ggtt->flags & XE_GGTT_FLAGS_64K)
 		align = max_t(u32, align, SZ_64K);
 
-	if (bo->ggtt_node && view->type == I915_GTT_VIEW_NORMAL) {
-		vma->node = bo->ggtt_node;
+	if (bo->ggtt_node[ggtt->tile->id] && view->type == I915_GTT_VIEW_NORMAL) {
+		vma->node = bo->ggtt_node[ggtt->tile->id];
 	} else if (view->type == I915_GTT_VIEW_NORMAL) {
 		u32 x, size = bo->ttm.base.size;
 
@@ -345,10 +345,12 @@ err:
 
 static void __xe_unpin_fb_vma(struct i915_vma *vma)
 {
+	u8 tile_id = vma->node->ggtt->tile->id;
+
 	if (vma->dpt)
 		xe_bo_unpin_map_no_vm(vma->dpt);
-	else if (!xe_ggtt_node_allocated(vma->bo->ggtt_node) ||
-		 vma->bo->ggtt_node->base.start != vma->node->base.start)
+	else if (!xe_ggtt_node_allocated(vma->bo->ggtt_node[tile_id]) ||
+		 vma->bo->ggtt_node[tile_id]->base.start != vma->node->base.start)
 		xe_ggtt_node_remove(vma->node, false);
 
 	ttm_bo_reserve(&vma->bo->ttm, false, false, NULL);
diff --git a/drivers/gpu/drm/xe/regs/xe_gt_regs.h b/drivers/gpu/drm/xe/regs/xe_gt_regs.h
index 0c9e4b2fafab..162f18e975da 100644
--- a/drivers/gpu/drm/xe/regs/xe_gt_regs.h
+++ b/drivers/gpu/drm/xe/regs/xe_gt_regs.h
@@ -445,6 +445,8 @@
 
 #define SAMPLER_MODE				XE_REG_MCR(0xe18c, XE_REG_OPTION_MASKED)
 #define   ENABLE_SMALLPL			REG_BIT(15)
+#define   SMP_WAIT_FETCH_MERGING_COUNTER	REG_GENMASK(11, 10)
+#define   SMP_FORCE_128B_OVERFETCH		REG_FIELD_PREP(SMP_WAIT_FETCH_MERGING_COUNTER, 1)
 #define   SC_DISABLE_POWER_OPTIMIZATION_EBB	REG_BIT(9)
 #define   SAMPLER_ENABLE_HEADLESS_MSG		REG_BIT(5)
 #define   INDIRECT_STATE_BASE_ADDR_OVERRIDE	REG_BIT(0)
diff --git a/drivers/gpu/drm/xe/regs/xe_oa_regs.h b/drivers/gpu/drm/xe/regs/xe_oa_regs.h
index a9b0091cb7ee..a49561e9f3c3 100644
--- a/drivers/gpu/drm/xe/regs/xe_oa_regs.h
+++ b/drivers/gpu/drm/xe/regs/xe_oa_regs.h
@@ -41,14 +41,6 @@
 
 #define OAG_OABUFFER		XE_REG(0xdb08)
 #define  OABUFFER_SIZE_MASK	REG_GENMASK(5, 3)
-#define  OABUFFER_SIZE_128K	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 0)
-#define  OABUFFER_SIZE_256K	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 1)
-#define  OABUFFER_SIZE_512K	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 2)
-#define  OABUFFER_SIZE_1M	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 3)
-#define  OABUFFER_SIZE_2M	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 4)
-#define  OABUFFER_SIZE_4M	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 5)
-#define  OABUFFER_SIZE_8M	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 6)
-#define  OABUFFER_SIZE_16M	REG_FIELD_PREP(OABUFFER_SIZE_MASK, 7)
 #define  OAG_OABUFFER_MEMORY_SELECT		REG_BIT(0) /* 0: PPGTT, 1: GGTT */
 
 #define OAG_OACONTROL				XE_REG(0xdaf4)
@@ -63,6 +55,7 @@
 #define OAG_OA_DEBUG XE_REG(0xdaf8, XE_REG_OPTION_MASKED)
 #define  OAG_OA_DEBUG_DISABLE_MMIO_TRG			REG_BIT(14)
 #define  OAG_OA_DEBUG_START_TRIGGER_SCOPE_CONTROL	REG_BIT(13)
+#define  OAG_OA_DEBUG_BUF_SIZE_SELECT			REG_BIT(12)
 #define  OAG_OA_DEBUG_DISABLE_START_TRG_2_COUNT_QUAL	REG_BIT(8)
 #define  OAG_OA_DEBUG_DISABLE_START_TRG_1_COUNT_QUAL	REG_BIT(7)
 #define  OAG_OA_DEBUG_INCLUDE_CLK_RATIO			REG_BIT(6)
diff --git a/drivers/gpu/drm/xe/regs/xe_pmt.h b/drivers/gpu/drm/xe/regs/xe_pmt.h
new file mode 100644
index 000000000000..f45abcd96ba8
--- /dev/null
+++ b/drivers/gpu/drm/xe/regs/xe_pmt.h
@@ -0,0 +1,19 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2024 Intel Corporation
+ */
+#ifndef _XE_PMT_H_
+#define _XE_PMT_H_
+
+#define SOC_BASE			0x280000
+
+#define BMG_PMT_BASE_OFFSET		0xDB000
+#define BMG_DISCOVERY_OFFSET		(SOC_BASE + BMG_PMT_BASE_OFFSET)
+
+#define BMG_TELEMETRY_BASE_OFFSET	0xE0000
+#define BMG_TELEMETRY_OFFSET		(SOC_BASE + BMG_TELEMETRY_BASE_OFFSET)
+
+#define SG_REMAP_INDEX1			XE_REG(SOC_BASE + 0x08)
+#define   SG_REMAP_BITS			REG_GENMASK(31, 24)
+
+#endif
diff --git a/drivers/gpu/drm/xe/tests/xe_bo.c b/drivers/gpu/drm/xe/tests/xe_bo.c
index 3e0ae40ebbd2..c9ec7a313c6b 100644
--- a/drivers/gpu/drm/xe/tests/xe_bo.c
+++ b/drivers/gpu/drm/xe/tests/xe_bo.c
@@ -49,6 +49,13 @@ static int ccs_test_migrate(struct xe_tile *tile, struct xe_bo *bo,
 			KUNIT_FAIL(test, "Failed to submit bo clear.\n");
 			return PTR_ERR(fence);
 		}
+
+		if (dma_fence_wait_timeout(fence, false, 5 * HZ) <= 0) {
+			dma_fence_put(fence);
+			KUNIT_FAIL(test, "Timeout while clearing bo.\n");
+			return  -ETIME;
+		}
+
 		dma_fence_put(fence);
 	}
 
diff --git a/drivers/gpu/drm/xe/tests/xe_migrate.c b/drivers/gpu/drm/xe/tests/xe_migrate.c
index 1a192a2a941b..d5fe0ea889ad 100644
--- a/drivers/gpu/drm/xe/tests/xe_migrate.c
+++ b/drivers/gpu/drm/xe/tests/xe_migrate.c
@@ -83,7 +83,8 @@ static void test_copy(struct xe_migrate *m, struct xe_bo *bo,
 						   bo->size,
 						   ttm_bo_type_kernel,
 						   region |
-						   XE_BO_FLAG_NEEDS_CPU_ACCESS);
+						   XE_BO_FLAG_NEEDS_CPU_ACCESS |
+						   XE_BO_FLAG_PINNED);
 	if (IS_ERR(remote)) {
 		KUNIT_FAIL(test, "Failed to allocate remote bo for %s: %pe\n",
 			   str, remote);
@@ -224,8 +225,8 @@ static void xe_migrate_sanity_test(struct xe_migrate *m, struct kunit *test)
 				    XE_BO_FLAG_VRAM_IF_DGFX(tile) |
 				    XE_BO_FLAG_PINNED);
 	if (IS_ERR(tiny)) {
-		KUNIT_FAIL(test, "Failed to allocate fake pt: %li\n",
-			   PTR_ERR(pt));
+		KUNIT_FAIL(test, "Failed to allocate tiny fake pt: %li\n",
+			   PTR_ERR(tiny));
 		goto free_pt;
 	}
 
@@ -642,7 +643,9 @@ static void validate_ccs_test_run_tile(struct xe_device *xe, struct xe_tile *til
 
 	sys_bo = xe_bo_create_user(xe, NULL, NULL, SZ_4M,
 				   DRM_XE_GEM_CPU_CACHING_WC,
-				   XE_BO_FLAG_SYSTEM | XE_BO_FLAG_NEEDS_CPU_ACCESS);
+				   XE_BO_FLAG_SYSTEM |
+				   XE_BO_FLAG_NEEDS_CPU_ACCESS |
+				   XE_BO_FLAG_PINNED);
 
 	if (IS_ERR(sys_bo)) {
 		KUNIT_FAIL(test, "xe_bo_create() failed with err=%ld\n",
@@ -666,7 +669,8 @@ static void validate_ccs_test_run_tile(struct xe_device *xe, struct xe_tile *til
 
 	ccs_bo = xe_bo_create_user(xe, NULL, NULL, SZ_4M,
 				   DRM_XE_GEM_CPU_CACHING_WC,
-				   bo_flags | XE_BO_FLAG_NEEDS_CPU_ACCESS);
+				   bo_flags | XE_BO_FLAG_NEEDS_CPU_ACCESS |
+				   XE_BO_FLAG_PINNED);
 
 	if (IS_ERR(ccs_bo)) {
 		KUNIT_FAIL(test, "xe_bo_create() failed with err=%ld\n",
@@ -690,7 +694,8 @@ static void validate_ccs_test_run_tile(struct xe_device *xe, struct xe_tile *til
 
 	vram_bo = xe_bo_create_user(xe, NULL, NULL, SZ_4M,
 				    DRM_XE_GEM_CPU_CACHING_WC,
-				    bo_flags | XE_BO_FLAG_NEEDS_CPU_ACCESS);
+				    bo_flags | XE_BO_FLAG_NEEDS_CPU_ACCESS |
+				    XE_BO_FLAG_PINNED);
 	if (IS_ERR(vram_bo)) {
 		KUNIT_FAIL(test, "xe_bo_create() failed with err=%ld\n",
 			   PTR_ERR(vram_bo));
diff --git a/drivers/gpu/drm/xe/xe_assert.h b/drivers/gpu/drm/xe/xe_assert.h
index 04d6b95c6d87..68fe70ce2be3 100644
--- a/drivers/gpu/drm/xe/xe_assert.h
+++ b/drivers/gpu/drm/xe/xe_assert.h
@@ -14,7 +14,7 @@
 #include "xe_step.h"
 
 /**
- * DOC: Xe ASSERTs
+ * DOC: Xe Asserts
  *
  * While Xe driver aims to be simpler than legacy i915 driver it is still
  * complex enough that some changes introduced while adding new functionality
@@ -103,7 +103,7 @@
  * (&CONFIG_DRM_XE_DEBUG must be enabled) and cannot be used in expressions
  * or as a condition.
  *
- * See `Xe ASSERTs`_ for general usage guidelines.
+ * See `Xe Asserts`_ for general usage guidelines.
  */
 #define xe_assert(xe, condition) xe_assert_msg((xe), condition, "")
 #define xe_assert_msg(xe, condition, msg, arg...) ({						\
@@ -138,7 +138,7 @@
  * (&CONFIG_DRM_XE_DEBUG must be enabled) and cannot be used in expressions
  * or as a condition.
  *
- * See `Xe ASSERTs`_ for general usage guidelines.
+ * See `Xe Asserts`_ for general usage guidelines.
  */
 #define xe_tile_assert(tile, condition) xe_tile_assert_msg((tile), condition, "")
 #define xe_tile_assert_msg(tile, condition, msg, arg...) ({					\
@@ -162,7 +162,7 @@
  * (&CONFIG_DRM_XE_DEBUG must be enabled) and cannot be used in expressions
  * or as a condition.
  *
- * See `Xe ASSERTs`_ for general usage guidelines.
+ * See `Xe Asserts`_ for general usage guidelines.
  */
 #define xe_gt_assert(gt, condition) xe_gt_assert_msg((gt), condition, "")
 #define xe_gt_assert_msg(gt, condition, msg, arg...) ({						\
diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c
index ae6b337cdc54..283cd0294570 100644
--- a/drivers/gpu/drm/xe/xe_bo.c
+++ b/drivers/gpu/drm/xe/xe_bo.c
@@ -162,6 +162,15 @@ static void try_add_system(struct xe_device *xe, struct xe_bo *bo,
 	}
 }
 
+static bool force_contiguous(u32 bo_flags)
+{
+	/*
+	 * For eviction / restore on suspend / resume objects pinned in VRAM
+	 * must be contiguous, also only contiguous BOs support xe_bo_vmap.
+	 */
+	return bo_flags & (XE_BO_FLAG_PINNED | XE_BO_FLAG_GGTT);
+}
+
 static void add_vram(struct xe_device *xe, struct xe_bo *bo,
 		     struct ttm_place *places, u32 bo_flags, u32 mem_type, u32 *c)
 {
@@ -175,12 +184,7 @@ static void add_vram(struct xe_device *xe, struct xe_bo *bo,
 	xe_assert(xe, vram && vram->usable_size);
 	io_size = vram->io_size;
 
-	/*
-	 * For eviction / restore on suspend / resume objects
-	 * pinned in VRAM must be contiguous
-	 */
-	if (bo_flags & (XE_BO_FLAG_PINNED |
-			XE_BO_FLAG_GGTT))
+	if (force_contiguous(bo_flags))
 		place.flags |= TTM_PL_FLAG_CONTIGUOUS;
 
 	if (io_size < vram->usable_size) {
@@ -212,8 +216,7 @@ static void try_add_stolen(struct xe_device *xe, struct xe_bo *bo,
 
 		bo->placements[*c] = (struct ttm_place) {
 			.mem_type = XE_PL_STOLEN,
-			.flags = bo_flags & (XE_BO_FLAG_PINNED |
-					     XE_BO_FLAG_GGTT) ?
+			.flags = force_contiguous(bo_flags) ?
 				TTM_PL_FLAG_CONTIGUOUS : 0,
 		};
 		*c += 1;
@@ -442,6 +445,14 @@ static void xe_ttm_tt_destroy(struct ttm_device *ttm_dev, struct ttm_tt *tt)
 	kfree(tt);
 }
 
+static bool xe_ttm_resource_visible(struct ttm_resource *mem)
+{
+	struct xe_ttm_vram_mgr_resource *vres =
+		to_xe_ttm_vram_mgr_resource(mem);
+
+	return vres->used_visible_size == mem->size;
+}
+
 static int xe_ttm_io_mem_reserve(struct ttm_device *bdev,
 				 struct ttm_resource *mem)
 {
@@ -453,11 +464,9 @@ static int xe_ttm_io_mem_reserve(struct ttm_device *bdev,
 		return 0;
 	case XE_PL_VRAM0:
 	case XE_PL_VRAM1: {
-		struct xe_ttm_vram_mgr_resource *vres =
-			to_xe_ttm_vram_mgr_resource(mem);
 		struct xe_mem_region *vram = res_to_mem_region(mem);
 
-		if (vres->used_visible_size < mem->size)
+		if (!xe_ttm_resource_visible(mem))
 			return -EINVAL;
 
 		mem->bus.offset = mem->start << PAGE_SHIFT;
@@ -876,6 +885,7 @@ int xe_bo_evict_pinned(struct xe_bo *bo)
 	};
 	struct ttm_operation_ctx ctx = {
 		.interruptible = false,
+		.gfp_retry_mayfail = true,
 	};
 	struct ttm_resource *new_mem;
 	int ret;
@@ -937,6 +947,7 @@ int xe_bo_restore_pinned(struct xe_bo *bo)
 {
 	struct ttm_operation_ctx ctx = {
 		.interruptible = false,
+		.gfp_retry_mayfail = false,
 	};
 	struct ttm_resource *new_mem;
 	struct ttm_place *place = &bo->placements[0];
@@ -1106,7 +1117,8 @@ static void xe_ttm_bo_purge(struct ttm_buffer_object *ttm_bo, struct ttm_operati
 static void xe_ttm_bo_swap_notify(struct ttm_buffer_object *ttm_bo)
 {
 	struct ttm_operation_ctx ctx = {
-		.interruptible = false
+		.interruptible = false,
+		.gfp_retry_mayfail = false,
 	};
 
 	if (ttm_bo->ttm) {
@@ -1118,6 +1130,52 @@ static void xe_ttm_bo_swap_notify(struct ttm_buffer_object *ttm_bo)
 	}
 }
 
+static int xe_ttm_access_memory(struct ttm_buffer_object *ttm_bo,
+				unsigned long offset, void *buf, int len,
+				int write)
+{
+	struct xe_bo *bo = ttm_to_xe_bo(ttm_bo);
+	struct xe_device *xe = ttm_to_xe_device(ttm_bo->bdev);
+	struct iosys_map vmap;
+	struct xe_res_cursor cursor;
+	struct xe_mem_region *vram;
+	int bytes_left = len;
+
+	xe_bo_assert_held(bo);
+	xe_device_assert_mem_access(xe);
+
+	if (!mem_type_is_vram(ttm_bo->resource->mem_type))
+		return -EIO;
+
+	/* FIXME: Use GPU for non-visible VRAM */
+	if (!xe_ttm_resource_visible(ttm_bo->resource))
+		return -EIO;
+
+	vram = res_to_mem_region(ttm_bo->resource);
+	xe_res_first(ttm_bo->resource, offset & PAGE_MASK,
+		     bo->size - (offset & PAGE_MASK), &cursor);
+
+	do {
+		unsigned long page_offset = (offset & ~PAGE_MASK);
+		int byte_count = min((int)(PAGE_SIZE - page_offset), bytes_left);
+
+		iosys_map_set_vaddr_iomem(&vmap, (u8 __iomem *)vram->mapping +
+					  cursor.start);
+		if (write)
+			xe_map_memcpy_to(xe, &vmap, page_offset, buf, byte_count);
+		else
+			xe_map_memcpy_from(xe, buf, &vmap, page_offset, byte_count);
+
+		buf += byte_count;
+		offset += byte_count;
+		bytes_left -= byte_count;
+		if (bytes_left)
+			xe_res_next(&cursor, PAGE_SIZE);
+	} while (bytes_left);
+
+	return len;
+}
+
 const struct ttm_device_funcs xe_ttm_funcs = {
 	.ttm_tt_create = xe_ttm_tt_create,
 	.ttm_tt_populate = xe_ttm_tt_populate,
@@ -1127,6 +1185,7 @@ const struct ttm_device_funcs xe_ttm_funcs = {
 	.move = xe_bo_move,
 	.io_mem_reserve = xe_ttm_io_mem_reserve,
 	.io_mem_pfn = xe_ttm_io_mem_pfn,
+	.access_memory = xe_ttm_access_memory,
 	.release_notify = xe_ttm_bo_release_notify,
 	.eviction_valuable = ttm_bo_eviction_valuable,
 	.delete_mem_notify = xe_ttm_bo_delete_mem_notify,
@@ -1137,6 +1196,8 @@ static void xe_ttm_bo_destroy(struct ttm_buffer_object *ttm_bo)
 {
 	struct xe_bo *bo = ttm_to_xe_bo(ttm_bo);
 	struct xe_device *xe = ttm_to_xe_device(ttm_bo->bdev);
+	struct xe_tile *tile;
+	u8 id;
 
 	if (bo->ttm.base.import_attach)
 		drm_prime_gem_destroy(&bo->ttm.base, NULL);
@@ -1144,8 +1205,9 @@ static void xe_ttm_bo_destroy(struct ttm_buffer_object *ttm_bo)
 
 	xe_assert(xe, list_empty(&ttm_bo->base.gpuva.list));
 
-	if (bo->ggtt_node && bo->ggtt_node->base.size)
-		xe_ggtt_remove_bo(bo->tile->mem.ggtt, bo);
+	for_each_tile(tile, xe, id)
+		if (bo->ggtt_node[id] && bo->ggtt_node[id]->base.size)
+			xe_ggtt_remove_bo(tile->mem.ggtt, bo);
 
 #ifdef CONFIG_PROC_FS
 	if (bo->client)
@@ -1243,11 +1305,50 @@ out:
 	return ret;
 }
 
+static int xe_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
+			   void *buf, int len, int write)
+{
+	struct ttm_buffer_object *ttm_bo = vma->vm_private_data;
+	struct xe_bo *bo = ttm_to_xe_bo(ttm_bo);
+	struct xe_device *xe = xe_bo_device(bo);
+	int ret;
+
+	xe_pm_runtime_get(xe);
+	ret = ttm_bo_vm_access(vma, addr, buf, len, write);
+	xe_pm_runtime_put(xe);
+
+	return ret;
+}
+
+/**
+ * xe_bo_read() - Read from an xe_bo
+ * @bo: The buffer object to read from.
+ * @offset: The byte offset to start reading from.
+ * @dst: Location to store the read.
+ * @size: Size in bytes for the read.
+ *
+ * Read @size bytes from the @bo, starting from @offset, storing into @dst.
+ *
+ * Return: Zero on success, or negative error.
+ */
+int xe_bo_read(struct xe_bo *bo, u64 offset, void *dst, int size)
+{
+	int ret;
+
+	ret = ttm_bo_access(&bo->ttm, offset, dst, size, 0);
+	if (ret >= 0 && ret != size)
+		ret = -EIO;
+	else if (ret == size)
+		ret = 0;
+
+	return ret;
+}
+
 static const struct vm_operations_struct xe_gem_vm_ops = {
 	.fault = xe_gem_fault,
 	.open = ttm_bo_vm_open,
 	.close = ttm_bo_vm_close,
-	.access = ttm_bo_vm_access
+	.access = xe_bo_vm_access,
 };
 
 static const struct drm_gem_object_funcs xe_gem_object_funcs = {
@@ -1301,6 +1402,7 @@ struct xe_bo *___xe_bo_create_locked(struct xe_device *xe, struct xe_bo *bo,
 	struct ttm_operation_ctx ctx = {
 		.interruptible = true,
 		.no_wait_gpu = false,
+		.gfp_retry_mayfail = true,
 	};
 	struct ttm_placement *placement;
 	uint32_t alignment;
@@ -1315,6 +1417,10 @@ struct xe_bo *___xe_bo_create_locked(struct xe_device *xe, struct xe_bo *bo,
 		return ERR_PTR(-EINVAL);
 	}
 
+	/* XE_BO_FLAG_GGTTx requires XE_BO_FLAG_GGTT also be set */
+	if ((flags & XE_BO_FLAG_GGTT_ALL) && !(flags & XE_BO_FLAG_GGTT))
+		return ERR_PTR(-EINVAL);
+
 	if (flags & (XE_BO_FLAG_VRAM_MASK | XE_BO_FLAG_STOLEN) &&
 	    !(flags & XE_BO_FLAG_IGNORE_MIN_PAGE_SIZE) &&
 	    ((xe->info.vram_flags & XE_VRAM_FLAGS_NEED64K) ||
@@ -1505,19 +1611,29 @@ __xe_bo_create_locked(struct xe_device *xe,
 	bo->vm = vm;
 
 	if (bo->flags & XE_BO_FLAG_GGTT) {
-		if (!tile && flags & XE_BO_FLAG_STOLEN)
-			tile = xe_device_get_root_tile(xe);
+		struct xe_tile *t;
+		u8 id;
 
-		xe_assert(xe, tile);
+		if (!(bo->flags & XE_BO_FLAG_GGTT_ALL)) {
+			if (!tile && flags & XE_BO_FLAG_STOLEN)
+				tile = xe_device_get_root_tile(xe);
 
-		if (flags & XE_BO_FLAG_FIXED_PLACEMENT) {
-			err = xe_ggtt_insert_bo_at(tile->mem.ggtt, bo,
-						   start + bo->size, U64_MAX);
-		} else {
-			err = xe_ggtt_insert_bo(tile->mem.ggtt, bo);
+			xe_assert(xe, tile);
+		}
+
+		for_each_tile(t, xe, id) {
+			if (t != tile && !(bo->flags & XE_BO_FLAG_GGTTx(t)))
+				continue;
+
+			if (flags & XE_BO_FLAG_FIXED_PLACEMENT) {
+				err = xe_ggtt_insert_bo_at(t->mem.ggtt, bo,
+							   start + bo->size, U64_MAX);
+			} else {
+				err = xe_ggtt_insert_bo(t->mem.ggtt, bo);
+			}
+			if (err)
+				goto err_unlock_put_bo;
 		}
-		if (err)
-			goto err_unlock_put_bo;
 	}
 
 	return bo;
@@ -1900,6 +2016,7 @@ int xe_bo_validate(struct xe_bo *bo, struct xe_vm *vm, bool allow_res_evict)
 	struct ttm_operation_ctx ctx = {
 		.interruptible = true,
 		.no_wait_gpu = false,
+		.gfp_retry_mayfail = true,
 	};
 
 	if (vm) {
@@ -1910,6 +2027,7 @@ int xe_bo_validate(struct xe_bo *bo, struct xe_vm *vm, bool allow_res_evict)
 		ctx.resv = xe_vm_resv(vm);
 	}
 
+	trace_xe_bo_validate(bo);
 	return ttm_bo_validate(&bo->ttm, &bo->placement, &ctx);
 }
 
@@ -1961,13 +2079,15 @@ dma_addr_t xe_bo_addr(struct xe_bo *bo, u64 offset, size_t page_size)
 
 int xe_bo_vmap(struct xe_bo *bo)
 {
+	struct xe_device *xe = ttm_to_xe_device(bo->ttm.bdev);
 	void *virtual;
 	bool is_iomem;
 	int ret;
 
 	xe_bo_assert_held(bo);
 
-	if (!(bo->flags & XE_BO_FLAG_NEEDS_CPU_ACCESS))
+	if (drm_WARN_ON(&xe->drm, !(bo->flags & XE_BO_FLAG_NEEDS_CPU_ACCESS) ||
+			!force_contiguous(bo->flags)))
 		return -EINVAL;
 
 	if (!iosys_map_is_null(&bo->vmap))
@@ -2243,6 +2363,7 @@ int xe_bo_migrate(struct xe_bo *bo, u32 mem_type)
 	struct ttm_operation_ctx ctx = {
 		.interruptible = true,
 		.no_wait_gpu = false,
+		.gfp_retry_mayfail = true,
 	};
 	struct ttm_placement placement;
 	struct ttm_place requested;
@@ -2293,6 +2414,7 @@ int xe_bo_evict(struct xe_bo *bo, bool force_alloc)
 		.interruptible = false,
 		.no_wait_gpu = false,
 		.force_alloc = force_alloc,
+		.gfp_retry_mayfail = true,
 	};
 	struct ttm_placement placement;
 	int ret;
@@ -2372,14 +2494,18 @@ void xe_bo_put_commit(struct llist_head *deferred)
 
 void xe_bo_put(struct xe_bo *bo)
 {
+	struct xe_tile *tile;
+	u8 id;
+
 	might_sleep();
 	if (bo) {
 #ifdef CONFIG_PROC_FS
 		if (bo->client)
 			might_lock(&bo->client->bos_lock);
 #endif
-		if (bo->ggtt_node && bo->ggtt_node->ggtt)
-			might_lock(&bo->ggtt_node->ggtt->lock);
+		for_each_tile(tile, xe_bo_device(bo), id)
+			if (bo->ggtt_node[id] && bo->ggtt_node[id]->ggtt)
+				might_lock(&bo->ggtt_node[id]->ggtt->lock);
 		drm_gem_object_put(&bo->ttm.base);
 	}
 }
diff --git a/drivers/gpu/drm/xe/xe_bo.h b/drivers/gpu/drm/xe/xe_bo.h
index 7fa44a0138b0..d9386ab03140 100644
--- a/drivers/gpu/drm/xe/xe_bo.h
+++ b/drivers/gpu/drm/xe/xe_bo.h
@@ -39,10 +39,22 @@
 #define XE_BO_FLAG_NEEDS_64K		BIT(15)
 #define XE_BO_FLAG_NEEDS_2M		BIT(16)
 #define XE_BO_FLAG_GGTT_INVALIDATE	BIT(17)
+#define XE_BO_FLAG_GGTT0                BIT(18)
+#define XE_BO_FLAG_GGTT1                BIT(19)
+#define XE_BO_FLAG_GGTT2                BIT(20)
+#define XE_BO_FLAG_GGTT3                BIT(21)
+#define XE_BO_FLAG_GGTT_ALL             (XE_BO_FLAG_GGTT0 | \
+					 XE_BO_FLAG_GGTT1 | \
+					 XE_BO_FLAG_GGTT2 | \
+					 XE_BO_FLAG_GGTT3)
+
 /* this one is trigger internally only */
 #define XE_BO_FLAG_INTERNAL_TEST	BIT(30)
 #define XE_BO_FLAG_INTERNAL_64K		BIT(31)
 
+#define XE_BO_FLAG_GGTTx(tile) \
+	(XE_BO_FLAG_GGTT0 << (tile)->id)
+
 #define XE_PTE_SHIFT			12
 #define XE_PAGE_SIZE			(1 << XE_PTE_SHIFT)
 #define XE_PTE_MASK			(XE_PAGE_SIZE - 1)
@@ -194,18 +206,29 @@ xe_bo_main_addr(struct xe_bo *bo, size_t page_size)
 }
 
 static inline u32
-xe_bo_ggtt_addr(struct xe_bo *bo)
+__xe_bo_ggtt_addr(struct xe_bo *bo, u8 tile_id)
 {
-	if (XE_WARN_ON(!bo->ggtt_node))
+	struct xe_ggtt_node *ggtt_node = bo->ggtt_node[tile_id];
+
+	if (XE_WARN_ON(!ggtt_node))
 		return 0;
 
-	XE_WARN_ON(bo->ggtt_node->base.size > bo->size);
-	XE_WARN_ON(bo->ggtt_node->base.start + bo->ggtt_node->base.size > (1ull << 32));
-	return bo->ggtt_node->base.start;
+	XE_WARN_ON(ggtt_node->base.size > bo->size);
+	XE_WARN_ON(ggtt_node->base.start + ggtt_node->base.size > (1ull << 32));
+	return ggtt_node->base.start;
+}
+
+static inline u32
+xe_bo_ggtt_addr(struct xe_bo *bo)
+{
+	xe_assert(xe_bo_device(bo), bo->tile);
+
+	return __xe_bo_ggtt_addr(bo, bo->tile->id);
 }
 
 int xe_bo_vmap(struct xe_bo *bo);
 void xe_bo_vunmap(struct xe_bo *bo);
+int xe_bo_read(struct xe_bo *bo, u64 offset, void *dst, int size);
 
 bool mem_type_is_vram(u32 mem_type);
 bool xe_bo_is_vram(struct xe_bo *bo);
diff --git a/drivers/gpu/drm/xe/xe_bo_evict.c b/drivers/gpu/drm/xe/xe_bo_evict.c
index 8fb2be061003..6a40eedd9db1 100644
--- a/drivers/gpu/drm/xe/xe_bo_evict.c
+++ b/drivers/gpu/drm/xe/xe_bo_evict.c
@@ -152,11 +152,17 @@ int xe_bo_restore_kernel(struct xe_device *xe)
 		}
 
 		if (bo->flags & XE_BO_FLAG_GGTT) {
-			struct xe_tile *tile = bo->tile;
+			struct xe_tile *tile;
+			u8 id;
 
-			mutex_lock(&tile->mem.ggtt->lock);
-			xe_ggtt_map_bo(tile->mem.ggtt, bo);
-			mutex_unlock(&tile->mem.ggtt->lock);
+			for_each_tile(tile, xe, id) {
+				if (tile != bo->tile && !(bo->flags & XE_BO_FLAG_GGTTx(tile)))
+					continue;
+
+				mutex_lock(&tile->mem.ggtt->lock);
+				xe_ggtt_map_bo(tile->mem.ggtt, bo);
+				mutex_unlock(&tile->mem.ggtt->lock);
+			}
 		}
 
 		/*
diff --git a/drivers/gpu/drm/xe/xe_bo_types.h b/drivers/gpu/drm/xe/xe_bo_types.h
index 99196228dcc8..46dc9e4e3e46 100644
--- a/drivers/gpu/drm/xe/xe_bo_types.h
+++ b/drivers/gpu/drm/xe/xe_bo_types.h
@@ -12,6 +12,7 @@
 #include <drm/ttm/ttm_device.h>
 #include <drm/ttm/ttm_placement.h>
 
+#include "xe_device_types.h"
 #include "xe_ggtt_types.h"
 
 struct xe_device;
@@ -38,8 +39,8 @@ struct xe_bo {
 	struct ttm_place placements[XE_BO_MAX_PLACEMENTS];
 	/** @placement: current placement for this BO */
 	struct ttm_placement placement;
-	/** @ggtt_node: GGTT node if this BO is mapped in the GGTT */
-	struct xe_ggtt_node *ggtt_node;
+	/** @ggtt_node: Array of GGTT nodes if this BO is mapped in the GGTTs */
+	struct xe_ggtt_node *ggtt_node[XE_MAX_TILES_PER_DEVICE];
 	/** @vmap: iosys map of this buffer */
 	struct iosys_map vmap;
 	/** @ttm_kmap: TTM bo kmap object for internal use only. Keep off. */
diff --git a/drivers/gpu/drm/xe/xe_devcoredump.c b/drivers/gpu/drm/xe/xe_devcoredump.c
index f8947e7e917e..71636e80b71d 100644
--- a/drivers/gpu/drm/xe/xe_devcoredump.c
+++ b/drivers/gpu/drm/xe/xe_devcoredump.c
@@ -30,30 +30,39 @@
 /**
  * DOC: Xe device coredump
  *
- * Devices overview:
  * Xe uses dev_coredump infrastructure for exposing the crash errors in a
- * standardized way.
- * devcoredump exposes a temporary device under /sys/class/devcoredump/
- * which is linked with our card device directly.
- * The core dump can be accessed either from
- * /sys/class/drm/card<n>/device/devcoredump/ or from
- * /sys/class/devcoredump/devcd<m> where
- * /sys/class/devcoredump/devcd<m>/failing_device is a link to
- * /sys/class/drm/card<n>/device/.
+ * standardized way. Once a crash occurs, devcoredump exposes a temporary
+ * node under ``/sys/class/devcoredump/devcd<m>/``. The same node is also
+ * accessible in ``/sys/class/drm/card<n>/device/devcoredump/``. The
+ * ``failing_device`` symlink points to the device that crashed and created the
+ * coredump.
  *
- * Snapshot at hang:
- * The 'data' file is printed with a drm_printer pointer at devcoredump read
- * time. For this reason, we need to take snapshots from when the hang has
- * happened, and not only when the user is reading the file. Otherwise the
- * information is outdated since the resets might have happened in between.
+ * The following characteristics are observed by xe when creating a device
+ * coredump:
  *
- * 'First' failure snapshot:
- * In general, the first hang is the most critical one since the following hangs
- * can be a consequence of the initial hang. For this reason we only take the
- * snapshot of the 'first' failure and ignore subsequent calls of this function,
- * at least while the coredump device is alive. Dev_coredump has a delayed work
- * queue that will eventually delete the device and free all the dump
- * information.
+ * **Snapshot at hang**:
+ *   The 'data' file contains a snapshot of the HW and driver states at the time
+ *   the hang happened. Due to the driver recovering from resets/crashes, it may
+ *   not correspond to the state of the system when the file is read by
+ *   userspace.
+ *
+ * **Coredump release**:
+ *   After a coredump is generated, it stays in kernel memory until released by
+ *   userpace by writing anything to it, or after an internal timer expires. The
+ *   exact timeout may vary and should not be relied upon. Example to release
+ *   a coredump:
+ *
+ *   .. code-block:: shell
+ *
+ *	$ > /sys/class/drm/card0/device/devcoredump/data
+ *
+ * **First failure only**:
+ *   In general, the first hang is the most critical one since the following
+ *   hangs can be a consequence of the initial hang. For this reason a snapshot
+ *   is taken only for the first failure. Until the devcoredump is released by
+ *   userspace or kernel, all subsequent hangs do not override the snapshot nor
+ *   create new ones. Devcoredump has a delayed work queue that will eventually
+ *   delete the file node and free all the dump information.
  */
 
 #ifdef CONFIG_DEV_COREDUMP
@@ -91,6 +100,7 @@ static ssize_t __xe_devcoredump_read(char *buffer, size_t count,
 	p = drm_coredump_printer(&iter);
 
 	drm_puts(&p, "**** Xe Device Coredump ****\n");
+	drm_printf(&p, "Reason: %s\n", ss->reason);
 	drm_puts(&p, "kernel: " UTS_RELEASE "\n");
 	drm_puts(&p, "module: " KBUILD_MODNAME "\n");
 
@@ -98,7 +108,7 @@ static ssize_t __xe_devcoredump_read(char *buffer, size_t count,
 	drm_printf(&p, "Snapshot time: %lld.%09ld\n", ts.tv_sec, ts.tv_nsec);
 	ts = ktime_to_timespec64(ss->boot_time);
 	drm_printf(&p, "Uptime: %lld.%09ld\n", ts.tv_sec, ts.tv_nsec);
-	drm_printf(&p, "Process: %s\n", ss->process_name);
+	drm_printf(&p, "Process: %s [%d]\n", ss->process_name, ss->pid);
 	xe_device_snapshot_print(xe, &p);
 
 	drm_printf(&p, "\n**** GT #%d ****\n", ss->gt->info.id);
@@ -130,6 +140,9 @@ static void xe_devcoredump_snapshot_free(struct xe_devcoredump_snapshot *ss)
 {
 	int i;
 
+	kfree(ss->reason);
+	ss->reason = NULL;
+
 	xe_guc_log_snapshot_free(ss->guc.log);
 	ss->guc.log = NULL;
 
@@ -170,16 +183,24 @@ static ssize_t xe_devcoredump_read(char *buffer, loff_t offset,
 	/* Ensure delayed work is captured before continuing */
 	flush_work(&ss->work);
 
-	if (!ss->read.buffer)
+	mutex_lock(&coredump->lock);
+
+	if (!ss->read.buffer) {
+		mutex_unlock(&coredump->lock);
 		return -ENODEV;
+	}
 
-	if (offset >= ss->read.size)
+	if (offset >= ss->read.size) {
+		mutex_unlock(&coredump->lock);
 		return 0;
+	}
 
 	byte_copied = count < ss->read.size - offset ? count :
 		ss->read.size - offset;
 	memcpy(buffer, ss->read.buffer + offset, byte_copied);
 
+	mutex_unlock(&coredump->lock);
+
 	return byte_copied;
 }
 
@@ -193,15 +214,18 @@ static void xe_devcoredump_free(void *data)
 
 	cancel_work_sync(&coredump->snapshot.work);
 
+	mutex_lock(&coredump->lock);
+
 	xe_devcoredump_snapshot_free(&coredump->snapshot);
 	kvfree(coredump->snapshot.read.buffer);
 
 	/* To prevent stale data on next snapshot, clear everything */
 	memset(&coredump->snapshot, 0, sizeof(coredump->snapshot));
 	coredump->captured = false;
-	coredump->job = NULL;
 	drm_info(&coredump_to_xe(coredump)->drm,
 		 "Xe device coredump has been deleted.\n");
+
+	mutex_unlock(&coredump->lock);
 }
 
 static void xe_devcoredump_deferred_snap_work(struct work_struct *work)
@@ -244,10 +268,10 @@ static void xe_devcoredump_deferred_snap_work(struct work_struct *work)
 }
 
 static void devcoredump_snapshot(struct xe_devcoredump *coredump,
+				 struct xe_exec_queue *q,
 				 struct xe_sched_job *job)
 {
 	struct xe_devcoredump_snapshot *ss = &coredump->snapshot;
-	struct xe_exec_queue *q = job->q;
 	struct xe_guc *guc = exec_queue_to_guc(q);
 	u32 adj_logical_mask = q->logical_mask;
 	u32 width_mask = (0x1 << q->width) - 1;
@@ -260,12 +284,14 @@ static void devcoredump_snapshot(struct xe_devcoredump *coredump,
 	ss->snapshot_time = ktime_get_real();
 	ss->boot_time = ktime_get_boottime();
 
-	if (q->vm && q->vm->xef)
+	if (q->vm && q->vm->xef) {
 		process_name = q->vm->xef->process_name;
+		ss->pid = q->vm->xef->pid;
+	}
+
 	strscpy(ss->process_name, process_name);
 
 	ss->gt = q->gt;
-	coredump->job = job;
 	INIT_WORK(&ss->work, xe_devcoredump_deferred_snap_work);
 
 	cookie = dma_fence_begin_signalling();
@@ -284,10 +310,11 @@ static void devcoredump_snapshot(struct xe_devcoredump *coredump,
 	ss->guc.log = xe_guc_log_snapshot_capture(&guc->log, true);
 	ss->guc.ct = xe_guc_ct_snapshot_capture(&guc->ct);
 	ss->ge = xe_guc_exec_queue_snapshot_capture(q);
-	ss->job = xe_sched_job_snapshot_capture(job);
+	if (job)
+		ss->job = xe_sched_job_snapshot_capture(job);
 	ss->vm = xe_vm_snapshot_capture(q->vm);
 
-	xe_engine_snapshot_capture_for_job(job);
+	xe_engine_snapshot_capture_for_queue(q);
 
 	queue_work(system_unbound_wq, &ss->work);
 
@@ -297,28 +324,42 @@ static void devcoredump_snapshot(struct xe_devcoredump *coredump,
 
 /**
  * xe_devcoredump - Take the required snapshots and initialize coredump device.
+ * @q: The faulty xe_exec_queue, where the issue was detected.
  * @job: The faulty xe_sched_job, where the issue was detected.
+ * @fmt: Printf format + args to describe the reason for the core dump
  *
  * This function should be called at the crash time within the serialized
  * gt_reset. It is skipped if we still have the core dump device available
  * with the information of the 'first' snapshot.
  */
-void xe_devcoredump(struct xe_sched_job *job)
+__printf(3, 4)
+void xe_devcoredump(struct xe_exec_queue *q, struct xe_sched_job *job, const char *fmt, ...)
 {
-	struct xe_device *xe = gt_to_xe(job->q->gt);
+	struct xe_device *xe = gt_to_xe(q->gt);
 	struct xe_devcoredump *coredump = &xe->devcoredump;
+	va_list varg;
+
+	mutex_lock(&coredump->lock);
 
 	if (coredump->captured) {
 		drm_dbg(&xe->drm, "Multiple hangs are occurring, but only the first snapshot was taken\n");
+		mutex_unlock(&coredump->lock);
 		return;
 	}
 
 	coredump->captured = true;
-	devcoredump_snapshot(coredump, job);
+
+	va_start(varg, fmt);
+	coredump->snapshot.reason = kvasprintf(GFP_ATOMIC, fmt, varg);
+	va_end(varg);
+
+	devcoredump_snapshot(coredump, q, job);
 
 	drm_info(&xe->drm, "Xe device coredump has been created\n");
 	drm_info(&xe->drm, "Check your /sys/class/drm/card%d/device/devcoredump/data\n",
 		 xe->drm.primary->index);
+
+	mutex_unlock(&coredump->lock);
 }
 
 static void xe_driver_devcoredump_fini(void *arg)
@@ -330,6 +371,18 @@ static void xe_driver_devcoredump_fini(void *arg)
 
 int xe_devcoredump_init(struct xe_device *xe)
 {
+	int err;
+
+	err = drmm_mutex_init(&xe->drm, &xe->devcoredump.lock);
+	if (err)
+		return err;
+
+	if (IS_ENABLED(CONFIG_LOCKDEP)) {
+		fs_reclaim_acquire(GFP_KERNEL);
+		might_lock(&xe->devcoredump.lock);
+		fs_reclaim_release(GFP_KERNEL);
+	}
+
 	return devm_add_action_or_reset(xe->drm.dev, xe_driver_devcoredump_fini, &xe->drm);
 }
 
diff --git a/drivers/gpu/drm/xe/xe_devcoredump.h b/drivers/gpu/drm/xe/xe_devcoredump.h
index a4eebc285fc8..6a17e6d60102 100644
--- a/drivers/gpu/drm/xe/xe_devcoredump.h
+++ b/drivers/gpu/drm/xe/xe_devcoredump.h
@@ -10,13 +10,16 @@
 
 struct drm_printer;
 struct xe_device;
+struct xe_exec_queue;
 struct xe_sched_job;
 
 #ifdef CONFIG_DEV_COREDUMP
-void xe_devcoredump(struct xe_sched_job *job);
+void xe_devcoredump(struct xe_exec_queue *q, struct xe_sched_job *job, const char *fmt, ...);
 int xe_devcoredump_init(struct xe_device *xe);
 #else
-static inline void xe_devcoredump(struct xe_sched_job *job)
+static inline void xe_devcoredump(struct xe_exec_queue *q,
+				  struct xe_sched_job *job,
+				  const char *fmt, ...)
 {
 }
 
diff --git a/drivers/gpu/drm/xe/xe_devcoredump_types.h b/drivers/gpu/drm/xe/xe_devcoredump_types.h
index 3703ddea1252..1a1d16a96b2d 100644
--- a/drivers/gpu/drm/xe/xe_devcoredump_types.h
+++ b/drivers/gpu/drm/xe/xe_devcoredump_types.h
@@ -28,6 +28,10 @@ struct xe_devcoredump_snapshot {
 	ktime_t boot_time;
 	/** @process_name: Name of process that triggered this gpu hang */
 	char process_name[TASK_COMM_LEN];
+	/** @pid: Process id of process that triggered this gpu hang */
+	pid_t pid;
+	/** @reason: The reason the coredump was triggered */
+	char *reason;
 
 	/** @gt: Affected GT, used by forcewake for delayed capture */
 	struct xe_gt *gt;
@@ -76,12 +80,12 @@ struct xe_devcoredump_snapshot {
  * for reading the information.
  */
 struct xe_devcoredump {
-	/** @captured: The snapshot of the first hang has already been taken. */
+	/** @lock: protects access to entire structure */
+	struct mutex lock;
+	/** @captured: The snapshot of the first hang has already been taken */
 	bool captured;
 	/** @snapshot: Snapshot is captured at time of the first crash */
 	struct xe_devcoredump_snapshot snapshot;
-	/** @job: Point to the faulting job */
-	struct xe_sched_job *job;
 };
 
 #endif
diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
index d8b6bc8fc8cf..56d4ffb650da 100644
--- a/drivers/gpu/drm/xe/xe_device.c
+++ b/drivers/gpu/drm/xe/xe_device.c
@@ -44,6 +44,7 @@
 #include "xe_memirq.h"
 #include "xe_mmio.h"
 #include "xe_module.h"
+#include "xe_oa.h"
 #include "xe_observation.h"
 #include "xe_pat.h"
 #include "xe_pcode.h"
@@ -55,6 +56,7 @@
 #include "xe_ttm_sys_mgr.h"
 #include "xe_vm.h"
 #include "xe_vram.h"
+#include "xe_vsec.h"
 #include "xe_wait_user_fence.h"
 #include "xe_wa.h"
 
@@ -365,6 +367,10 @@ struct xe_device *xe_device_create(struct pci_dev *pdev,
 		goto err;
 	}
 
+	err = drmm_mutex_init(&xe->drm, &xe->pmt.lock);
+	if (err)
+		goto err;
+
 	err = xe_display_create(xe);
 	if (WARN_ON(err))
 		goto err;
@@ -759,6 +765,8 @@ int xe_device_probe(struct xe_device *xe)
 	for_each_gt(gt, xe, id)
 		xe_gt_sanitize_freq(gt);
 
+	xe_vsec_init(xe);
+
 	return devm_add_action_or_reset(xe->drm.dev, xe_device_sanitize, xe);
 
 err_fini_display:
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index b9ea455d6f59..ace22e35e769 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -16,7 +16,7 @@
 #include "xe_heci_gsc.h"
 #include "xe_lmtt_types.h"
 #include "xe_memirq_types.h"
-#include "xe_oa.h"
+#include "xe_oa_types.h"
 #include "xe_platform_types.h"
 #include "xe_pt_types.h"
 #include "xe_sriov_types.h"
@@ -42,8 +42,6 @@ struct xe_pat_ops;
 #define GRAPHICS_VERx100(xe) ((xe)->info.graphics_verx100)
 #define MEDIA_VERx100(xe) ((xe)->info.media_verx100)
 #define IS_DGFX(xe) ((xe)->info.is_dgfx)
-#define HAS_HECI_GSCFI(xe) ((xe)->info.has_heci_gscfi)
-#define HAS_HECI_CSCFI(xe) ((xe)->info.has_heci_cscfi)
 
 #define XE_VRAM_FLAGS_NEED64K		BIT(0)
 
@@ -296,14 +294,24 @@ struct xe_device {
 		/** @info.va_bits: Maximum bits of a virtual address */
 		u8 va_bits;
 
-		/** @info.is_dgfx: is discrete device */
-		u8 is_dgfx:1;
-		/** @info.has_asid: Has address space ID */
-		u8 has_asid:1;
+		/*
+		 * Keep all flags below alphabetically sorted
+		 */
+
 		/** @info.force_execlist: Forced execlist submission */
 		u8 force_execlist:1;
+		/** @info.has_asid: Has address space ID */
+		u8 has_asid:1;
+		/** @info.has_atomic_enable_pte_bit: Device has atomic enable PTE bit */
+		u8 has_atomic_enable_pte_bit:1;
+		/** @info.has_device_atomics_on_smem: Supports device atomics on SMEM */
+		u8 has_device_atomics_on_smem:1;
 		/** @info.has_flat_ccs: Whether flat CCS metadata is used */
 		u8 has_flat_ccs:1;
+		/** @info.has_heci_cscfi: device has heci cscfi */
+		u8 has_heci_cscfi:1;
+		/** @info.has_heci_gscfi: device has heci gscfi */
+		u8 has_heci_gscfi:1;
 		/** @info.has_llc: Device has a shared CPU+GPU last level cache */
 		u8 has_llc:1;
 		/** @info.has_mmio_ext: Device has extra MMIO address range */
@@ -314,6 +322,8 @@ struct xe_device {
 		u8 has_sriov:1;
 		/** @info.has_usm: Device has unified shared memory support */
 		u8 has_usm:1;
+		/** @info.is_dgfx: is discrete device */
+		u8 is_dgfx:1;
 		/**
 		 * @info.probe_display: Probe display hardware.  If set to
 		 * false, the driver will behave as if there is no display
@@ -323,20 +333,12 @@ struct xe_device {
 		 * state the firmware or bootloader left it in.
 		 */
 		u8 probe_display:1;
+		/** @info.skip_guc_pc: Skip GuC based PM feature init */
+		u8 skip_guc_pc:1;
 		/** @info.skip_mtcfg: skip Multi-Tile configuration from MTCFG register */
 		u8 skip_mtcfg:1;
 		/** @info.skip_pcode: skip access to PCODE uC */
 		u8 skip_pcode:1;
-		/** @info.has_heci_gscfi: device has heci gscfi */
-		u8 has_heci_gscfi:1;
-		/** @info.has_heci_cscfi: device has heci cscfi */
-		u8 has_heci_cscfi:1;
-		/** @info.skip_guc_pc: Skip GuC based PM feature init */
-		u8 skip_guc_pc:1;
-		/** @info.has_atomic_enable_pte_bit: Device has atomic enable PTE bit */
-		u8 has_atomic_enable_pte_bit:1;
-		/** @info.has_device_atomics_on_smem: Supports device atomics on SMEM */
-		u8 has_device_atomics_on_smem:1;
 	} info;
 
 	/** @irq: device interrupt state */
@@ -345,7 +347,7 @@ struct xe_device {
 		spinlock_t lock;
 
 		/** @irq.enabled: interrupts enabled on this device */
-		bool enabled;
+		atomic_t enabled;
 	} irq;
 
 	/** @ttm: ttm device */
@@ -374,6 +376,8 @@ struct xe_device {
 
 		/** @sriov.pf: PF specific data */
 		struct xe_device_pf pf;
+		/** @sriov.vf: VF specific data */
+		struct xe_device_vf vf;
 
 		/** @sriov.wq: workqueue used by the virtualization workers */
 		struct workqueue_struct *wq;
@@ -481,6 +485,12 @@ struct xe_device {
 		struct mutex lock;
 	} d3cold;
 
+	/** @pmt: Support the PMT driver callback interface */
+	struct {
+		/** @pmt.lock: protect access for telemetry data */
+		struct mutex lock;
+	} pmt;
+
 	/**
 	 * @pm_callback_task: Track the active task that is running in either
 	 * the runtime_suspend or runtime_resume callbacks.
@@ -588,7 +598,7 @@ struct xe_file {
 		/** @vm.xe: xarray to store VMs */
 		struct xarray xa;
 		/**
-		 * @vm.lock: Protects VM lookup + reference and removal a from
+		 * @vm.lock: Protects VM lookup + reference and removal from
 		 * file xarray. Not an intended to be an outer lock which does
 		 * thing while being held.
 		 */
@@ -601,10 +611,15 @@ struct xe_file {
 		struct xarray xa;
 		/**
 		 * @exec_queue.lock: Protects exec queue lookup + reference and
-		 * removal a frommfile xarray. Not an intended to be an outer
-		 * lock which does thing while being held.
+		 * removal from file xarray. Not intended to be an outer lock
+		 * which does things while being held.
 		 */
 		struct mutex lock;
+		/**
+		 * @exec_queue.pending_removal: items pending to be removed to
+		 * synchronize GPU state update with ongoing query.
+		 */
+		atomic_t pending_removal;
 	} exec_queue;
 
 	/** @run_ticks: hw engine class run time in ticks for this drm client */
diff --git a/drivers/gpu/drm/xe/xe_drm_client.c b/drivers/gpu/drm/xe/xe_drm_client.c
index 22f0f1a6dfd5..298a587da7f1 100644
--- a/drivers/gpu/drm/xe/xe_drm_client.c
+++ b/drivers/gpu/drm/xe/xe_drm_client.c
@@ -269,6 +269,49 @@ static void show_meminfo(struct drm_printer *p, struct drm_file *file)
 	}
 }
 
+static struct xe_hw_engine *any_engine(struct xe_device *xe)
+{
+	struct xe_gt *gt;
+	unsigned long gt_id;
+
+	for_each_gt(gt, xe, gt_id) {
+		struct xe_hw_engine *hwe = xe_gt_any_hw_engine(gt);
+
+		if (hwe)
+			return hwe;
+	}
+
+	return NULL;
+}
+
+static bool force_wake_get_any_engine(struct xe_device *xe,
+				      struct xe_hw_engine **phwe,
+				      unsigned int *pfw_ref)
+{
+	enum xe_force_wake_domains domain;
+	unsigned int fw_ref;
+	struct xe_hw_engine *hwe;
+	struct xe_force_wake *fw;
+
+	hwe = any_engine(xe);
+	if (!hwe)
+		return false;
+
+	domain = xe_hw_engine_to_fw_domain(hwe);
+	fw = gt_to_fw(hwe->gt);
+
+	fw_ref = xe_force_wake_get(fw, domain);
+	if (!xe_force_wake_ref_has_domain(fw_ref, domain)) {
+		xe_force_wake_put(fw, fw_ref);
+		return false;
+	}
+
+	*phwe = hwe;
+	*pfw_ref = fw_ref;
+
+	return true;
+}
+
 static void show_run_ticks(struct drm_printer *p, struct drm_file *file)
 {
 	unsigned long class, i, gt_id, capacity[XE_ENGINE_CLASS_MAX] = { };
@@ -280,7 +323,18 @@ static void show_run_ticks(struct drm_printer *p, struct drm_file *file)
 	u64 gpu_timestamp;
 	unsigned int fw_ref;
 
+	/*
+	 * Wait for any exec queue going away: their cycles will get updated on
+	 * context switch out, so wait for that to happen
+	 */
+	wait_var_event(&xef->exec_queue.pending_removal,
+		       !atomic_read(&xef->exec_queue.pending_removal));
+
 	xe_pm_runtime_get(xe);
+	if (!force_wake_get_any_engine(xe, &hwe, &fw_ref)) {
+		xe_pm_runtime_put(xe);
+		return;
+	}
 
 	/* Accumulate all the exec queues from this client */
 	mutex_lock(&xef->exec_queue.lock);
@@ -295,33 +349,11 @@ static void show_run_ticks(struct drm_printer *p, struct drm_file *file)
 	}
 	mutex_unlock(&xef->exec_queue.lock);
 
-	/* Get the total GPU cycles */
-	for_each_gt(gt, xe, gt_id) {
-		enum xe_force_wake_domains fw;
-
-		hwe = xe_gt_any_hw_engine(gt);
-		if (!hwe)
-			continue;
-
-		fw = xe_hw_engine_to_fw_domain(hwe);
-
-		fw_ref = xe_force_wake_get(gt_to_fw(gt), fw);
-		if (!xe_force_wake_ref_has_domain(fw_ref, fw)) {
-			hwe = NULL;
-			xe_force_wake_put(gt_to_fw(gt), fw_ref);
-			break;
-		}
-
-		gpu_timestamp = xe_hw_engine_read_timestamp(hwe);
-		xe_force_wake_put(gt_to_fw(gt), fw_ref);
-		break;
-	}
+	gpu_timestamp = xe_hw_engine_read_timestamp(hwe);
 
+	xe_force_wake_put(gt_to_fw(hwe->gt), fw_ref);
 	xe_pm_runtime_put(xe);
 
-	if (unlikely(!hwe))
-		return;
-
 	for (class = 0; class < XE_ENGINE_CLASS_MAX; class++) {
 		const char *class_name;
 
diff --git a/drivers/gpu/drm/xe/xe_exec_queue.c b/drivers/gpu/drm/xe/xe_exec_queue.c
index fd0f3b3c9101..aab9e561153d 100644
--- a/drivers/gpu/drm/xe/xe_exec_queue.c
+++ b/drivers/gpu/drm/xe/xe_exec_queue.c
@@ -240,6 +240,7 @@ struct xe_exec_queue *xe_exec_queue_create_bind(struct xe_device *xe,
 
 	return q;
 }
+ALLOW_ERROR_INJECTION(xe_exec_queue_create_bind, ERRNO);
 
 void xe_exec_queue_destroy(struct kref *ref)
 {
@@ -262,8 +263,11 @@ void xe_exec_queue_fini(struct xe_exec_queue *q)
 
 	/*
 	 * Before releasing our ref to lrc and xef, accumulate our run ticks
+	 * and wakeup any waiters.
 	 */
 	xe_exec_queue_update_run_ticks(q);
+	if (q->xef && atomic_dec_and_test(&q->xef->exec_queue.pending_removal))
+		wake_up_var(&q->xef->exec_queue.pending_removal);
 
 	for (i = 0; i < q->width; ++i)
 		xe_lrc_put(q->lrc[i]);
@@ -826,7 +830,10 @@ int xe_exec_queue_destroy_ioctl(struct drm_device *dev, void *data,
 
 	mutex_lock(&xef->exec_queue.lock);
 	q = xa_erase(&xef->exec_queue.xa, args->exec_queue_id);
+	if (q)
+		atomic_inc(&xef->exec_queue.pending_removal);
 	mutex_unlock(&xef->exec_queue.lock);
+
 	if (XE_IOCTL_DBG(xe, !q))
 		return -ENOENT;
 
diff --git a/drivers/gpu/drm/xe/xe_ggtt.c b/drivers/gpu/drm/xe/xe_ggtt.c
index 558fac8bb6fb..05154f9de1a6 100644
--- a/drivers/gpu/drm/xe/xe_ggtt.c
+++ b/drivers/gpu/drm/xe/xe_ggtt.c
@@ -598,10 +598,10 @@ void xe_ggtt_map_bo(struct xe_ggtt *ggtt, struct xe_bo *bo)
 	u64 start;
 	u64 offset, pte;
 
-	if (XE_WARN_ON(!bo->ggtt_node))
+	if (XE_WARN_ON(!bo->ggtt_node[ggtt->tile->id]))
 		return;
 
-	start = bo->ggtt_node->base.start;
+	start = bo->ggtt_node[ggtt->tile->id]->base.start;
 
 	for (offset = 0; offset < bo->size; offset += XE_PAGE_SIZE) {
 		pte = ggtt->pt_ops->pte_encode_bo(bo, offset, pat_index);
@@ -612,15 +612,16 @@ void xe_ggtt_map_bo(struct xe_ggtt *ggtt, struct xe_bo *bo)
 static int __xe_ggtt_insert_bo_at(struct xe_ggtt *ggtt, struct xe_bo *bo,
 				  u64 start, u64 end)
 {
-	int err;
 	u64 alignment = bo->min_align > 0 ? bo->min_align : XE_PAGE_SIZE;
+	u8 tile_id = ggtt->tile->id;
+	int err;
 
 	if (xe_bo_is_vram(bo) && ggtt->flags & XE_GGTT_FLAGS_64K)
 		alignment = SZ_64K;
 
-	if (XE_WARN_ON(bo->ggtt_node)) {
+	if (XE_WARN_ON(bo->ggtt_node[tile_id])) {
 		/* Someone's already inserted this BO in the GGTT */
-		xe_tile_assert(ggtt->tile, bo->ggtt_node->base.size == bo->size);
+		xe_tile_assert(ggtt->tile, bo->ggtt_node[tile_id]->base.size == bo->size);
 		return 0;
 	}
 
@@ -630,19 +631,19 @@ static int __xe_ggtt_insert_bo_at(struct xe_ggtt *ggtt, struct xe_bo *bo,
 
 	xe_pm_runtime_get_noresume(tile_to_xe(ggtt->tile));
 
-	bo->ggtt_node = xe_ggtt_node_init(ggtt);
-	if (IS_ERR(bo->ggtt_node)) {
-		err = PTR_ERR(bo->ggtt_node);
-		bo->ggtt_node = NULL;
+	bo->ggtt_node[tile_id] = xe_ggtt_node_init(ggtt);
+	if (IS_ERR(bo->ggtt_node[tile_id])) {
+		err = PTR_ERR(bo->ggtt_node[tile_id]);
+		bo->ggtt_node[tile_id] = NULL;
 		goto out;
 	}
 
 	mutex_lock(&ggtt->lock);
-	err = drm_mm_insert_node_in_range(&ggtt->mm, &bo->ggtt_node->base, bo->size,
-					  alignment, 0, start, end, 0);
+	err = drm_mm_insert_node_in_range(&ggtt->mm, &bo->ggtt_node[tile_id]->base,
+					  bo->size, alignment, 0, start, end, 0);
 	if (err) {
-		xe_ggtt_node_fini(bo->ggtt_node);
-		bo->ggtt_node = NULL;
+		xe_ggtt_node_fini(bo->ggtt_node[tile_id]);
+		bo->ggtt_node[tile_id] = NULL;
 	} else {
 		xe_ggtt_map_bo(ggtt, bo);
 	}
@@ -691,13 +692,15 @@ int xe_ggtt_insert_bo(struct xe_ggtt *ggtt, struct xe_bo *bo)
  */
 void xe_ggtt_remove_bo(struct xe_ggtt *ggtt, struct xe_bo *bo)
 {
-	if (XE_WARN_ON(!bo->ggtt_node))
+	u8 tile_id = ggtt->tile->id;
+
+	if (XE_WARN_ON(!bo->ggtt_node[tile_id]))
 		return;
 
 	/* This BO is not currently in the GGTT */
-	xe_tile_assert(ggtt->tile, bo->ggtt_node->base.size == bo->size);
+	xe_tile_assert(ggtt->tile, bo->ggtt_node[tile_id]->base.size == bo->size);
 
-	xe_ggtt_node_remove(bo->ggtt_node,
+	xe_ggtt_node_remove(bo->ggtt_node[tile_id],
 			    bo->flags & XE_BO_FLAG_GGTT_INVALIDATE);
 }
 
diff --git a/drivers/gpu/drm/xe/xe_gpu_scheduler.h b/drivers/gpu/drm/xe/xe_gpu_scheduler.h
index 64b2ae6839db..c250ea773491 100644
--- a/drivers/gpu/drm/xe/xe_gpu_scheduler.h
+++ b/drivers/gpu/drm/xe/xe_gpu_scheduler.h
@@ -71,8 +71,14 @@ static inline void xe_sched_add_pending_job(struct xe_gpu_scheduler *sched,
 static inline
 struct xe_sched_job *xe_sched_first_pending_job(struct xe_gpu_scheduler *sched)
 {
-	return list_first_entry_or_null(&sched->base.pending_list,
-					struct xe_sched_job, drm.list);
+	struct xe_sched_job *job;
+
+	spin_lock(&sched->base.job_list_lock);
+	job = list_first_entry_or_null(&sched->base.pending_list,
+				       struct xe_sched_job, drm.list);
+	spin_unlock(&sched->base.job_list_lock);
+
+	return job;
 }
 
 static inline int
diff --git a/drivers/gpu/drm/xe/xe_gsc_proxy.c b/drivers/gpu/drm/xe/xe_gsc_proxy.c
index fc64b45d324b..24cc6a4f9a96 100644
--- a/drivers/gpu/drm/xe/xe_gsc_proxy.c
+++ b/drivers/gpu/drm/xe/xe_gsc_proxy.c
@@ -139,17 +139,29 @@ static int proxy_send_to_gsc(struct xe_gsc *gsc, u32 size)
 	return 0;
 }
 
-static int validate_proxy_header(struct xe_gsc_proxy_header *header,
+static int validate_proxy_header(struct xe_gt *gt,
+				 struct xe_gsc_proxy_header *header,
 				 u32 source, u32 dest, u32 max_size)
 {
 	u32 type = FIELD_GET(GSC_PROXY_TYPE, header->hdr);
 	u32 length = FIELD_GET(GSC_PROXY_PAYLOAD_LENGTH, header->hdr);
+	int ret = 0;
 
-	if (header->destination != dest || header->source != source)
-		return -ENOEXEC;
+	if (header->destination != dest || header->source != source) {
+		ret = -ENOEXEC;
+		goto out;
+	}
 
-	if (length + PROXY_HDR_SIZE > max_size)
-		return -E2BIG;
+	if (length + PROXY_HDR_SIZE > max_size) {
+		ret = -E2BIG;
+		goto out;
+	}
+
+	/* We only care about the status if this is a message for the driver */
+	if (dest == GSC_PROXY_ADDRESSING_KMD && header->status != 0) {
+		ret = -EIO;
+		goto out;
+	}
 
 	switch (type) {
 	case GSC_PROXY_MSG_TYPE_PROXY_PAYLOAD:
@@ -157,12 +169,20 @@ static int validate_proxy_header(struct xe_gsc_proxy_header *header,
 			break;
 		fallthrough;
 	case GSC_PROXY_MSG_TYPE_PROXY_INVALID:
-		return -EIO;
+		ret = -EIO;
+		break;
 	default:
 		break;
 	}
 
-	return 0;
+out:
+	if (ret)
+		xe_gt_err(gt,
+			  "GSC proxy error: s=0x%x[0x%x], d=0x%x[0x%x], t=%u, l=0x%x, st=0x%x\n",
+			  header->source, source, header->destination, dest,
+			  type, length, header->status);
+
+	return ret;
 }
 
 #define proxy_header_wr(xe_, map_, offset_, field_, val_) \
@@ -228,12 +248,17 @@ static int proxy_query(struct xe_gsc *gsc)
 		xe_map_memcpy_from(xe, to_csme_hdr, &gsc->proxy.from_gsc,
 				   reply_offset, PROXY_HDR_SIZE);
 
-		/* stop if this was the last message */
-		if (FIELD_GET(GSC_PROXY_TYPE, to_csme_hdr->hdr) == GSC_PROXY_MSG_TYPE_PROXY_END)
+		/* Check the status and stop if this was the last message */
+		if (FIELD_GET(GSC_PROXY_TYPE, to_csme_hdr->hdr) == GSC_PROXY_MSG_TYPE_PROXY_END) {
+			ret = validate_proxy_header(gt, to_csme_hdr,
+						    GSC_PROXY_ADDRESSING_GSC,
+						    GSC_PROXY_ADDRESSING_KMD,
+						    GSC_PROXY_BUFFER_SIZE - reply_offset);
 			break;
+		}
 
 		/* make sure the GSC-to-CSME proxy header is sane */
-		ret = validate_proxy_header(to_csme_hdr,
+		ret = validate_proxy_header(gt, to_csme_hdr,
 					    GSC_PROXY_ADDRESSING_GSC,
 					    GSC_PROXY_ADDRESSING_CSME,
 					    GSC_PROXY_BUFFER_SIZE - reply_offset);
@@ -262,7 +287,7 @@ static int proxy_query(struct xe_gsc *gsc)
 		}
 
 		/* make sure the CSME-to-GSC proxy header is sane */
-		ret = validate_proxy_header(gsc->proxy.from_csme,
+		ret = validate_proxy_header(gt, gsc->proxy.from_csme,
 					    GSC_PROXY_ADDRESSING_CSME,
 					    GSC_PROXY_ADDRESSING_GSC,
 					    GSC_PROXY_BUFFER_SIZE - reply_offset);
diff --git a/drivers/gpu/drm/xe/xe_gt.c b/drivers/gpu/drm/xe/xe_gt.c
index d6744be01a68..41ab7fbebc19 100644
--- a/drivers/gpu/drm/xe/xe_gt.c
+++ b/drivers/gpu/drm/xe/xe_gt.c
@@ -748,10 +748,8 @@ static int do_gt_restart(struct xe_gt *gt)
 	if (err)
 		return err;
 
-	for_each_hw_engine(hwe, gt, id) {
+	for_each_hw_engine(hwe, gt, id)
 		xe_reg_sr_apply_mmio(&hwe->reg_sr, gt);
-		xe_reg_sr_apply_whitelist(hwe);
-	}
 
 	/* Get CCS mode in sync between sw/hw */
 	xe_gt_apply_ccs_mode(gt);
diff --git a/drivers/gpu/drm/xe/xe_gt_printk.h b/drivers/gpu/drm/xe/xe_gt_printk.h
index 5dc71394372d..11da0228cea7 100644
--- a/drivers/gpu/drm/xe/xe_gt_printk.h
+++ b/drivers/gpu/drm/xe/xe_gt_printk.h
@@ -60,6 +60,21 @@ static inline void __xe_gt_printfn_info(struct drm_printer *p, struct va_format
 	xe_gt_info(gt, "%pV", vaf);
 }
 
+static inline void __xe_gt_printfn_dbg(struct drm_printer *p, struct va_format *vaf)
+{
+	struct xe_gt *gt = p->arg;
+	struct drm_printer dbg;
+
+	/*
+	 * The original xe_gt_dbg() callsite annotations are useless here,
+	 * redirect to the tweaked drm_dbg_printer() instead.
+	 */
+	dbg = drm_dbg_printer(&gt_to_xe(gt)->drm, DRM_UT_DRIVER, NULL);
+	dbg.origin = p->origin;
+
+	drm_printf(&dbg, "GT%u: %pV", gt->info.id, vaf);
+}
+
 /**
  * xe_gt_err_printer - Construct a &drm_printer that outputs to xe_gt_err()
  * @gt: the &xe_gt pointer to use in xe_gt_err()
@@ -90,4 +105,20 @@ static inline struct drm_printer xe_gt_info_printer(struct xe_gt *gt)
 	return p;
 }
 
+/**
+ * xe_gt_dbg_printer - Construct a &drm_printer that outputs like xe_gt_dbg()
+ * @gt: the &xe_gt pointer to use in xe_gt_dbg()
+ *
+ * Return: The &drm_printer object.
+ */
+static inline struct drm_printer xe_gt_dbg_printer(struct xe_gt *gt)
+{
+	struct drm_printer p = {
+		.printfn = __xe_gt_printfn_dbg,
+		.arg = gt,
+		.origin = (const void *)_THIS_IP_,
+	};
+	return p;
+}
+
 #endif
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c
index 192643d63d22..65082f12f1a8 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c
@@ -207,6 +207,11 @@ static int pf_push_vf_cfg_preempt_timeout(struct xe_gt *gt, unsigned int vfid, u
 	return pf_push_vf_cfg_u32(gt, vfid, GUC_KLV_VF_CFG_PREEMPT_TIMEOUT_KEY, *preempt_timeout);
 }
 
+static int pf_push_vf_cfg_sched_priority(struct xe_gt *gt, unsigned int vfid, u32 priority)
+{
+	return pf_push_vf_cfg_u32(gt, vfid, GUC_KLV_VF_CFG_SCHED_PRIORITY_KEY, priority);
+}
+
 static int pf_push_vf_cfg_lmem(struct xe_gt *gt, unsigned int vfid, u64 size)
 {
 	return pf_push_vf_cfg_u64(gt, vfid, GUC_KLV_VF_CFG_LMEM_SIZE_KEY, size);
@@ -1540,8 +1545,6 @@ static u64 pf_query_max_lmem(struct xe_gt *gt)
 
 #ifdef CONFIG_DRM_XE_DEBUG_SRIOV
 #define MAX_FAIR_LMEM	SZ_128M	/* XXX: make it small for the driver bringup */
-#else
-#define MAX_FAIR_LMEM	SZ_2G	/* XXX: known issue with allocating BO over 2GiB */
 #endif
 
 static u64 pf_estimate_fair_lmem(struct xe_gt *gt, unsigned int num_vfs)
@@ -1767,6 +1770,77 @@ u32 xe_gt_sriov_pf_config_get_preempt_timeout(struct xe_gt *gt, unsigned int vfi
 	return preempt_timeout;
 }
 
+static const char *sched_priority_unit(u32 priority)
+{
+	return priority == GUC_SCHED_PRIORITY_LOW ? "(low)" :
+		priority == GUC_SCHED_PRIORITY_NORMAL ? "(normal)" :
+		priority == GUC_SCHED_PRIORITY_HIGH ? "(high)" :
+		"(?)";
+}
+
+static int pf_provision_sched_priority(struct xe_gt *gt, unsigned int vfid, u32 priority)
+{
+	struct xe_gt_sriov_config *config = pf_pick_vf_config(gt, vfid);
+	int err;
+
+	err = pf_push_vf_cfg_sched_priority(gt, vfid, priority);
+	if (unlikely(err))
+		return err;
+
+	config->sched_priority = priority;
+	return 0;
+}
+
+static int pf_get_sched_priority(struct xe_gt *gt, unsigned int vfid)
+{
+	struct xe_gt_sriov_config *config = pf_pick_vf_config(gt, vfid);
+
+	return config->sched_priority;
+}
+
+/**
+ * xe_gt_sriov_pf_config_set_sched_priority() - Configure scheduling priority.
+ * @gt: the &xe_gt
+ * @vfid: the VF identifier
+ * @priority: requested scheduling priority
+ *
+ * This function can only be called on PF.
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int xe_gt_sriov_pf_config_set_sched_priority(struct xe_gt *gt, unsigned int vfid, u32 priority)
+{
+	int err;
+
+	mutex_lock(xe_gt_sriov_pf_master_mutex(gt));
+	err = pf_provision_sched_priority(gt, vfid, priority);
+	mutex_unlock(xe_gt_sriov_pf_master_mutex(gt));
+
+	return pf_config_set_u32_done(gt, vfid, priority,
+				      xe_gt_sriov_pf_config_get_sched_priority(gt, vfid),
+				      "scheduling priority", sched_priority_unit, err);
+}
+
+/**
+ * xe_gt_sriov_pf_config_get_sched_priority - Get VF's scheduling priority.
+ * @gt: the &xe_gt
+ * @vfid: the VF identifier
+ *
+ * This function can only be called on PF.
+ *
+ * Return: VF's (or PF's) scheduling priority.
+ */
+u32 xe_gt_sriov_pf_config_get_sched_priority(struct xe_gt *gt, unsigned int vfid)
+{
+	u32 priority;
+
+	mutex_lock(xe_gt_sriov_pf_master_mutex(gt));
+	priority = pf_get_sched_priority(gt, vfid);
+	mutex_unlock(xe_gt_sriov_pf_master_mutex(gt));
+
+	return priority;
+}
+
 static void pf_reset_config_sched(struct xe_gt *gt, struct xe_gt_sriov_config *config)
 {
 	lockdep_assert_held(xe_gt_sriov_pf_master_mutex(gt));
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.h b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.h
index 0c55aa40a1a7..f894e9d4abba 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.h
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.h
@@ -44,6 +44,9 @@ u32 xe_gt_sriov_pf_config_get_preempt_timeout(struct xe_gt *gt, unsigned int vfi
 int xe_gt_sriov_pf_config_set_preempt_timeout(struct xe_gt *gt, unsigned int vfid,
 					      u32 preempt_timeout);
 
+u32 xe_gt_sriov_pf_config_get_sched_priority(struct xe_gt *gt, unsigned int vfid);
+int xe_gt_sriov_pf_config_set_sched_priority(struct xe_gt *gt, unsigned int vfid, u32 priority);
+
 u32 xe_gt_sriov_pf_config_get_threshold(struct xe_gt *gt, unsigned int vfid,
 					enum xe_guc_klv_threshold_index index);
 int xe_gt_sriov_pf_config_set_threshold(struct xe_gt *gt, unsigned int vfid,
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config_types.h b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config_types.h
index 2d3b73d78f14..686c7b3b6d7a 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config_types.h
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config_types.h
@@ -33,6 +33,8 @@ struct xe_gt_sriov_config {
 	u32 exec_quantum;
 	/** @preempt_timeout: preemption timeout in microseconds. */
 	u32 preempt_timeout;
+	/** @sched_priority: scheduling priority. */
+	u32 sched_priority;
 	/** @thresholds: GuC thresholds for adverse events notifications. */
 	u32 thresholds[XE_GUC_KLV_NUM_THRESHOLDS];
 };
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c
index 05df4ab3514b..b2521dd6ec42 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_debugfs.c
@@ -164,6 +164,7 @@ static void pf_add_policy_attrs(struct xe_gt *gt, struct dentry *parent)
  *      │   │   ├── contexts_spare
  *      │   │   ├── exec_quantum_ms
  *      │   │   ├── preempt_timeout_us
+ *      │   │   ├── sched_priority
  *      │   ├── vf1
  *      │   │   ├── ggtt_quota
  *      │   │   ├── lmem_quota
@@ -171,6 +172,7 @@ static void pf_add_policy_attrs(struct xe_gt *gt, struct dentry *parent)
  *      │   │   ├── contexts_quota
  *      │   │   ├── exec_quantum_ms
  *      │   │   ├── preempt_timeout_us
+ *      │   │   ├── sched_priority
  */
 
 #define DEFINE_SRIOV_GT_CONFIG_DEBUGFS_ATTRIBUTE(CONFIG, TYPE, FORMAT)		\
@@ -209,6 +211,7 @@ DEFINE_SRIOV_GT_CONFIG_DEBUGFS_ATTRIBUTE(ctxs, u32, "%llu\n");
 DEFINE_SRIOV_GT_CONFIG_DEBUGFS_ATTRIBUTE(dbs, u32, "%llu\n");
 DEFINE_SRIOV_GT_CONFIG_DEBUGFS_ATTRIBUTE(exec_quantum, u32, "%llu\n");
 DEFINE_SRIOV_GT_CONFIG_DEBUGFS_ATTRIBUTE(preempt_timeout, u32, "%llu\n");
+DEFINE_SRIOV_GT_CONFIG_DEBUGFS_ATTRIBUTE(sched_priority, u32, "%llu\n");
 
 /*
  *      /sys/kernel/debug/dri/0/
@@ -295,6 +298,8 @@ static void pf_add_config_attrs(struct xe_gt *gt, struct dentry *parent, unsigne
 				   &exec_quantum_fops);
 	debugfs_create_file_unsafe("preempt_timeout_us", 0644, parent, parent,
 				   &preempt_timeout_fops);
+	debugfs_create_file_unsafe("sched_priority", 0644, parent, parent,
+				   &sched_priority_fops);
 
 	/* register all threshold attributes */
 #define register_threshold_attribute(TAG, NAME, ...) \
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_helpers.h b/drivers/gpu/drm/xe/xe_gt_sriov_pf_helpers.h
index 0bf12d89ceb2..6af219d93c3b 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_helpers.h
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_helpers.h
@@ -18,7 +18,7 @@
  * is within a range of supported VF numbers (up to maximum number of VFs that
  * driver can support, including VF0 that represents the PF itself).
  *
- * Note: Effective only on debug builds. See `Xe ASSERTs`_ for more information.
+ * Note: Effective only on debug builds. See `Xe Asserts`_ for more information.
  */
 #define xe_gt_sriov_pf_assert_vfid(gt, vfid)	xe_sriov_pf_assert_vfid(gt_to_xe(gt), (vfid))
 
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_policy.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_policy.c
index fae5be5a2a11..c00fb354705f 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_policy.c
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_policy.c
@@ -135,14 +135,33 @@ static int pf_update_policy_u32(struct xe_gt *gt, u16 key, u32 *policy, u32 valu
 	return 0;
 }
 
+static void pf_bulk_reset_sched_priority(struct xe_gt *gt, u32 priority)
+{
+	unsigned int total_vfs = 1 + xe_gt_sriov_pf_get_totalvfs(gt);
+	unsigned int n;
+
+	xe_gt_assert(gt, IS_SRIOV_PF(gt_to_xe(gt)));
+	lockdep_assert_held(xe_gt_sriov_pf_master_mutex(gt));
+
+	for (n = 0; n < total_vfs; n++)
+		gt->sriov.pf.vfs[n].config.sched_priority = priority;
+}
+
 static int pf_provision_sched_if_idle(struct xe_gt *gt, bool enable)
 {
+	int err;
+
 	xe_gt_assert(gt, IS_SRIOV_PF(gt_to_xe(gt)));
 	lockdep_assert_held(xe_gt_sriov_pf_master_mutex(gt));
 
-	return pf_update_policy_bool(gt, GUC_KLV_VGT_POLICY_SCHED_IF_IDLE_KEY,
-				     &gt->sriov.pf.policy.guc.sched_if_idle,
-				     enable);
+	err = pf_update_policy_bool(gt, GUC_KLV_VGT_POLICY_SCHED_IF_IDLE_KEY,
+				    &gt->sriov.pf.policy.guc.sched_if_idle,
+				    enable);
+
+	if (!err)
+		pf_bulk_reset_sched_priority(gt, enable ? GUC_SCHED_PRIORITY_NORMAL :
+					     GUC_SCHED_PRIORITY_LOW);
+	return err;
 }
 
 static int pf_reprovision_sched_if_idle(struct xe_gt *gt)
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_vf.c b/drivers/gpu/drm/xe/xe_gt_sriov_vf.c
index d3baba50f085..cca5d5732802 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_vf.c
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_vf.c
@@ -27,6 +27,7 @@
 #include "xe_guc_relay.h"
 #include "xe_mmio.h"
 #include "xe_sriov.h"
+#include "xe_sriov_vf.h"
 #include "xe_uc_fw.h"
 #include "xe_wopcm.h"
 
@@ -223,6 +224,44 @@ int xe_gt_sriov_vf_bootstrap(struct xe_gt *gt)
 	return 0;
 }
 
+static int guc_action_vf_notify_resfix_done(struct xe_guc *guc)
+{
+	u32 request[GUC_HXG_REQUEST_MSG_MIN_LEN] = {
+		FIELD_PREP(GUC_HXG_MSG_0_ORIGIN, GUC_HXG_ORIGIN_HOST) |
+		FIELD_PREP(GUC_HXG_MSG_0_TYPE, GUC_HXG_TYPE_REQUEST) |
+		FIELD_PREP(GUC_HXG_REQUEST_MSG_0_ACTION, GUC_ACTION_VF2GUC_NOTIFY_RESFIX_DONE),
+	};
+	int ret;
+
+	ret = xe_guc_mmio_send(guc, request, ARRAY_SIZE(request));
+
+	return ret > 0 ? -EPROTO : ret;
+}
+
+/**
+ * xe_gt_sriov_vf_notify_resfix_done - Notify GuC about resource fixups apply completed.
+ * @gt: the &xe_gt struct instance linked to target GuC
+ *
+ * Returns: 0 if the operation completed successfully, or a negative error
+ * code otherwise.
+ */
+int xe_gt_sriov_vf_notify_resfix_done(struct xe_gt *gt)
+{
+	struct xe_guc *guc = &gt->uc.guc;
+	int err;
+
+	xe_gt_assert(gt, IS_SRIOV_VF(gt_to_xe(gt)));
+
+	err = guc_action_vf_notify_resfix_done(guc);
+	if (unlikely(err))
+		xe_gt_sriov_err(gt, "Failed to notify GuC about resource fixup done (%pe)\n",
+				ERR_PTR(err));
+	else
+		xe_gt_sriov_dbg_verbose(gt, "sent GuC resource fixup done\n");
+
+	return err;
+}
+
 static int guc_action_query_single_klv(struct xe_guc *guc, u32 key,
 				       u32 *value, u32 value_len)
 {
@@ -692,6 +731,30 @@ failed:
 	return err;
 }
 
+/**
+ * xe_gt_sriov_vf_migrated_event_handler - Start a VF migration recovery,
+ *   or just mark that a GuC is ready for it.
+ * @gt: the &xe_gt struct instance linked to target GuC
+ *
+ * This function shall be called only by VF.
+ */
+void xe_gt_sriov_vf_migrated_event_handler(struct xe_gt *gt)
+{
+	struct xe_device *xe = gt_to_xe(gt);
+
+	xe_gt_assert(gt, IS_SRIOV_VF(xe));
+
+	set_bit(gt->info.id, &xe->sriov.vf.migration.gt_flags);
+	/*
+	 * We need to be certain that if all flags were set, at least one
+	 * thread will notice that and schedule the recovery.
+	 */
+	smp_mb__after_atomic();
+
+	xe_gt_sriov_info(gt, "ready for recovery after migration\n");
+	xe_sriov_vf_start_migration_recovery(xe);
+}
+
 static bool vf_is_negotiated(struct xe_gt *gt, u16 major, u16 minor)
 {
 	xe_gt_assert(gt, IS_SRIOV_VF(gt_to_xe(gt)));
diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_vf.h b/drivers/gpu/drm/xe/xe_gt_sriov_vf.h
index e541ce57bec2..912d20814261 100644
--- a/drivers/gpu/drm/xe/xe_gt_sriov_vf.h
+++ b/drivers/gpu/drm/xe/xe_gt_sriov_vf.h
@@ -17,6 +17,8 @@ int xe_gt_sriov_vf_query_config(struct xe_gt *gt);
 int xe_gt_sriov_vf_connect(struct xe_gt *gt);
 int xe_gt_sriov_vf_query_runtime(struct xe_gt *gt);
 int xe_gt_sriov_vf_prepare_ggtt(struct xe_gt *gt);
+int xe_gt_sriov_vf_notify_resfix_done(struct xe_gt *gt);
+void xe_gt_sriov_vf_migrated_event_handler(struct xe_gt *gt);
 
 u32 xe_gt_sriov_vf_gmdid(struct xe_gt *gt);
 u16 xe_gt_sriov_vf_guc_ids(struct xe_gt *gt);
diff --git a/drivers/gpu/drm/xe/xe_gt_stats.h b/drivers/gpu/drm/xe/xe_gt_stats.h
index 91d944f6c4e4..38325ef53617 100644
--- a/drivers/gpu/drm/xe/xe_gt_stats.h
+++ b/drivers/gpu/drm/xe/xe_gt_stats.h
@@ -6,15 +6,11 @@
 #ifndef _XE_GT_STATS_H_
 #define _XE_GT_STATS_H_
 
+#include "xe_gt_stats_types.h"
+
 struct xe_gt;
 struct drm_printer;
 
-enum xe_gt_stats_id {
-	XE_GT_STATS_ID_TLB_INVAL,
-	/* must be the last entry */
-	__XE_GT_STATS_NUM_IDS,
-};
-
 #ifdef CONFIG_DEBUG_FS
 int xe_gt_stats_print_info(struct xe_gt *gt, struct drm_printer *p);
 void xe_gt_stats_incr(struct xe_gt *gt, const enum xe_gt_stats_id id, int incr);
diff --git a/drivers/gpu/drm/xe/xe_gt_stats_types.h b/drivers/gpu/drm/xe/xe_gt_stats_types.h
new file mode 100644
index 000000000000..2fc055e39f27
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_gt_stats_types.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2024 Intel Corporation
+ */
+
+#ifndef _XE_GT_STATS_TYPES_H_
+#define _XE_GT_STATS_TYPES_H_
+
+enum xe_gt_stats_id {
+	XE_GT_STATS_ID_TLB_INVAL,
+	/* must be the last entry */
+	__XE_GT_STATS_NUM_IDS,
+};
+
+#endif
diff --git a/drivers/gpu/drm/xe/xe_gt_throttle.c b/drivers/gpu/drm/xe/xe_gt_throttle.c
index 03b225364101..8db78d616b6f 100644
--- a/drivers/gpu/drm/xe/xe_gt_throttle.c
+++ b/drivers/gpu/drm/xe/xe_gt_throttle.c
@@ -8,6 +8,7 @@
 #include <regs/xe_gt_regs.h>
 #include "xe_device.h"
 #include "xe_gt.h"
+#include "xe_gt_printk.h"
 #include "xe_gt_sysfs.h"
 #include "xe_gt_throttle.h"
 #include "xe_mmio.h"
@@ -53,6 +54,7 @@ static u32 read_status(struct xe_gt *gt)
 {
 	u32 status = xe_gt_throttle_get_limit_reasons(gt) & GT0_PERF_LIMIT_REASONS_MASK;
 
+	xe_gt_dbg(gt, "throttle reasons: 0x%08x\n", status);
 	return status;
 }
 
diff --git a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c
index 3cb228c773cd..665927b80e9e 100644
--- a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c
+++ b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c
@@ -65,6 +65,14 @@ invalidation_fence_signal(struct xe_device *xe, struct xe_gt_tlb_invalidation_fe
 	__invalidation_fence_signal(xe, fence);
 }
 
+void xe_gt_tlb_invalidation_fence_signal(struct xe_gt_tlb_invalidation_fence *fence)
+{
+	if (WARN_ON_ONCE(!fence->gt))
+		return;
+
+	__invalidation_fence_signal(gt_to_xe(fence->gt), fence);
+}
+
 static void xe_gt_tlb_fence_timeout(struct work_struct *work)
 {
 	struct xe_gt *gt = container_of(work, struct xe_gt,
@@ -253,9 +261,18 @@ static int xe_gt_tlb_invalidation_guc(struct xe_gt *gt,
 		0,  /* seqno, replaced in send_tlb_invalidation */
 		MAKE_INVAL_OP(XE_GUC_TLB_INVAL_GUC),
 	};
+	int ret;
+
+	ret = send_tlb_invalidation(&gt->uc.guc, fence, action,
+				    ARRAY_SIZE(action));
+	/*
+	 * -ECANCELED indicates the CT is stopped for a GT reset. TLB caches
+	 *  should be nuked on a GT reset so this error can be ignored.
+	 */
+	if (ret == -ECANCELED)
+		return 0;
 
-	return send_tlb_invalidation(&gt->uc.guc, fence, action,
-				     ARRAY_SIZE(action));
+	return ret;
 }
 
 /**
diff --git a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.h b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.h
index f430d5797af7..00b1c6c01e8d 100644
--- a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.h
+++ b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.h
@@ -28,6 +28,7 @@ int xe_guc_tlb_invalidation_done_handler(struct xe_guc *guc, u32 *msg, u32 len);
 void xe_gt_tlb_invalidation_fence_init(struct xe_gt *gt,
 				       struct xe_gt_tlb_invalidation_fence *fence,
 				       bool stack);
+void xe_gt_tlb_invalidation_fence_signal(struct xe_gt_tlb_invalidation_fence *fence);
 
 static inline void
 xe_gt_tlb_invalidation_fence_wait(struct xe_gt_tlb_invalidation_fence *fence)
diff --git a/drivers/gpu/drm/xe/xe_gt_types.h b/drivers/gpu/drm/xe/xe_gt_types.h
index a287b98ee70b..6e66bf0e8b3f 100644
--- a/drivers/gpu/drm/xe/xe_gt_types.h
+++ b/drivers/gpu/drm/xe/xe_gt_types.h
@@ -11,10 +11,10 @@
 #include "xe_gt_idle_types.h"
 #include "xe_gt_sriov_pf_types.h"
 #include "xe_gt_sriov_vf_types.h"
-#include "xe_gt_stats.h"
+#include "xe_gt_stats_types.h"
 #include "xe_hw_engine_types.h"
 #include "xe_hw_fence_types.h"
-#include "xe_oa.h"
+#include "xe_oa_types.h"
 #include "xe_reg_sr_types.h"
 #include "xe_sa_types.h"
 #include "xe_uc_types.h"
diff --git a/drivers/gpu/drm/xe/xe_guc.c b/drivers/gpu/drm/xe/xe_guc.c
index 7f704346a8f4..4e2868efb620 100644
--- a/drivers/gpu/drm/xe/xe_guc.c
+++ b/drivers/gpu/drm/xe/xe_guc.c
@@ -44,7 +44,15 @@ static u32 guc_bo_ggtt_addr(struct xe_guc *guc,
 			    struct xe_bo *bo)
 {
 	struct xe_device *xe = guc_to_xe(guc);
-	u32 addr = xe_bo_ggtt_addr(bo);
+	u32 addr;
+
+	/*
+	 * For most BOs, the address on the allocating tile is fine. However for
+	 * some, e.g. G2G CTB, the address on a specific tile is required as it
+	 * might be different for each tile. So, just always ask for the address
+	 * on the target GuC.
+	 */
+	addr = __xe_bo_ggtt_addr(bo, gt_to_tile(guc_to_gt(guc))->id);
 
 	/* GuC addresses above GUC_GGTT_TOP don't map through the GTT */
 	xe_assert(xe, addr >= xe_wopcm_size(guc_to_xe(guc)));
@@ -244,6 +252,293 @@ static void guc_write_params(struct xe_guc *guc)
 		xe_mmio_write32(&gt->mmio, SOFT_SCRATCH(1 + i), guc->params[i]);
 }
 
+static int guc_action_register_g2g_buffer(struct xe_guc *guc, u32 type, u32 dst_tile, u32 dst_dev,
+					  u32 desc_addr, u32 buff_addr, u32 size)
+{
+	struct xe_gt *gt = guc_to_gt(guc);
+	struct xe_device *xe = gt_to_xe(gt);
+	u32 action[] = {
+		XE_GUC_ACTION_REGISTER_G2G,
+		FIELD_PREP(XE_G2G_REGISTER_SIZE, size / SZ_4K - 1) |
+		FIELD_PREP(XE_G2G_REGISTER_TYPE, type) |
+		FIELD_PREP(XE_G2G_REGISTER_TILE, dst_tile) |
+		FIELD_PREP(XE_G2G_REGISTER_DEVICE, dst_dev),
+		desc_addr,
+		buff_addr,
+	};
+
+	xe_assert(xe, (type == XE_G2G_TYPE_IN) || (type == XE_G2G_TYPE_OUT));
+	xe_assert(xe, !(size % SZ_4K));
+
+	return xe_guc_ct_send_block(&guc->ct, action, ARRAY_SIZE(action));
+}
+
+static int guc_action_deregister_g2g_buffer(struct xe_guc *guc, u32 type, u32 dst_tile, u32 dst_dev)
+{
+	struct xe_gt *gt = guc_to_gt(guc);
+	struct xe_device *xe = gt_to_xe(gt);
+	u32 action[] = {
+		XE_GUC_ACTION_DEREGISTER_G2G,
+		FIELD_PREP(XE_G2G_DEREGISTER_TYPE, type) |
+		FIELD_PREP(XE_G2G_DEREGISTER_TILE, dst_tile) |
+		FIELD_PREP(XE_G2G_DEREGISTER_DEVICE, dst_dev),
+	};
+
+	xe_assert(xe, (type == XE_G2G_TYPE_IN) || (type == XE_G2G_TYPE_OUT));
+
+	return xe_guc_ct_send_block(&guc->ct, action, ARRAY_SIZE(action));
+}
+
+#define G2G_DEV(gt)	(((gt)->info.type == XE_GT_TYPE_MAIN) ? 0 : 1)
+
+#define G2G_BUFFER_SIZE (SZ_4K)
+#define G2G_DESC_SIZE (64)
+#define G2G_DESC_AREA_SIZE (SZ_4K)
+
+/*
+ * Generate a unique id for each bi-directional CTB for each pair of
+ * near and far tiles/devices. The id can then be used as an index into
+ * a single allocation that is sub-divided into multiple CTBs.
+ *
+ * For example, with two devices per tile and two tiles, the table should
+ * look like:
+ *           Far <tile>.<dev>
+ *         0.0   0.1   1.0   1.1
+ * N 0.0  --/-- 00/01 02/03 04/05
+ * e 0.1  01/00 --/-- 06/07 08/09
+ * a 1.0  03/02 07/06 --/-- 10/11
+ * r 1.1  05/04 09/08 11/10 --/--
+ *
+ * Where each entry is Rx/Tx channel id.
+ *
+ * So GuC #3 (tile 1, dev 1) talking to GuC #2 (tile 1, dev 0) would
+ * be reading from channel #11 and writing to channel #10. Whereas,
+ * GuC #2 talking to GuC #3 would be read on #10 and write to #11.
+ */
+static unsigned int g2g_slot(u32 near_tile, u32 near_dev, u32 far_tile, u32 far_dev,
+			     u32 type, u32 max_inst, bool have_dev)
+{
+	u32 near = near_tile, far = far_tile;
+	u32 idx = 0, x, y, direction;
+	int i;
+
+	if (have_dev) {
+		near = (near << 1) | near_dev;
+		far = (far << 1) | far_dev;
+	}
+
+	/* No need to send to one's self */
+	if (far == near)
+		return -1;
+
+	if (far > near) {
+		/* Top right table half */
+		x = far;
+		y = near;
+
+		/* T/R is 'forwards' direction */
+		direction = type;
+	} else {
+		/* Bottom left table half */
+		x = near;
+		y = far;
+
+		/* B/L is 'backwards' direction */
+		direction = (1 - type);
+	}
+
+	/* Count the rows prior to the target */
+	for (i = y; i > 0; i--)
+		idx += max_inst - i;
+
+	/* Count this row up to the target */
+	idx += (x - 1 - y);
+
+	/* Slots are in Rx/Tx pairs */
+	idx *= 2;
+
+	/* Pick Rx/Tx direction */
+	idx += direction;
+
+	return idx;
+}
+
+static int guc_g2g_register(struct xe_guc *near_guc, struct xe_gt *far_gt, u32 type, bool have_dev)
+{
+	struct xe_gt *near_gt = guc_to_gt(near_guc);
+	struct xe_device *xe = gt_to_xe(near_gt);
+	struct xe_bo *g2g_bo;
+	u32 near_tile = gt_to_tile(near_gt)->id;
+	u32 near_dev = G2G_DEV(near_gt);
+	u32 far_tile = gt_to_tile(far_gt)->id;
+	u32 far_dev = G2G_DEV(far_gt);
+	u32 max = xe->info.gt_count;
+	u32 base, desc, buf;
+	int slot;
+
+	/* G2G is not allowed between different cards */
+	xe_assert(xe, xe == gt_to_xe(far_gt));
+
+	g2g_bo = near_guc->g2g.bo;
+	xe_assert(xe, g2g_bo);
+
+	slot = g2g_slot(near_tile, near_dev, far_tile, far_dev, type, max, have_dev);
+	xe_assert(xe, slot >= 0);
+
+	base = guc_bo_ggtt_addr(near_guc, g2g_bo);
+	desc = base + slot * G2G_DESC_SIZE;
+	buf = base + G2G_DESC_AREA_SIZE + slot * G2G_BUFFER_SIZE;
+
+	xe_assert(xe, (desc - base + G2G_DESC_SIZE) <= G2G_DESC_AREA_SIZE);
+	xe_assert(xe, (buf - base + G2G_BUFFER_SIZE) <= g2g_bo->size);
+
+	return guc_action_register_g2g_buffer(near_guc, type, far_tile, far_dev,
+					      desc, buf, G2G_BUFFER_SIZE);
+}
+
+static void guc_g2g_deregister(struct xe_guc *guc, u32 far_tile, u32 far_dev, u32 type)
+{
+	guc_action_deregister_g2g_buffer(guc, type, far_tile, far_dev);
+}
+
+static u32 guc_g2g_size(struct xe_guc *guc)
+{
+	struct xe_gt *gt = guc_to_gt(guc);
+	struct xe_device *xe = gt_to_xe(gt);
+	unsigned int count = xe->info.gt_count;
+	u32 num_channels = (count * (count - 1)) / 2;
+
+	xe_assert(xe, num_channels * XE_G2G_TYPE_LIMIT * G2G_DESC_SIZE <= G2G_DESC_AREA_SIZE);
+
+	return num_channels * XE_G2G_TYPE_LIMIT * G2G_BUFFER_SIZE + G2G_DESC_AREA_SIZE;
+}
+
+static bool xe_guc_g2g_wanted(struct xe_device *xe)
+{
+	/* Can't do GuC to GuC communication if there is only one GuC */
+	if (xe->info.gt_count <= 1)
+		return false;
+
+	/* No current user */
+	return false;
+}
+
+static int guc_g2g_alloc(struct xe_guc *guc)
+{
+	struct xe_gt *gt = guc_to_gt(guc);
+	struct xe_device *xe = gt_to_xe(gt);
+	struct xe_tile *tile = gt_to_tile(gt);
+	struct xe_bo *bo;
+	u32 g2g_size;
+
+	if (guc->g2g.bo)
+		return 0;
+
+	if (gt->info.id != 0) {
+		struct xe_gt *root_gt = xe_device_get_gt(xe, 0);
+		struct xe_guc *root_guc = &root_gt->uc.guc;
+		struct xe_bo *bo;
+
+		bo = xe_bo_get(root_guc->g2g.bo);
+		if (!bo)
+			return -ENODEV;
+
+		guc->g2g.bo = bo;
+		guc->g2g.owned = false;
+		return 0;
+	}
+
+	g2g_size = guc_g2g_size(guc);
+	bo = xe_managed_bo_create_pin_map(xe, tile, g2g_size,
+					  XE_BO_FLAG_VRAM_IF_DGFX(tile) |
+					  XE_BO_FLAG_GGTT |
+					  XE_BO_FLAG_GGTT_ALL |
+					  XE_BO_FLAG_GGTT_INVALIDATE);
+	if (IS_ERR(bo))
+		return PTR_ERR(bo);
+
+	xe_map_memset(xe, &bo->vmap, 0, 0, g2g_size);
+	guc->g2g.bo = bo;
+	guc->g2g.owned = true;
+
+	return 0;
+}
+
+static void guc_g2g_fini(struct xe_guc *guc)
+{
+	if (!guc->g2g.bo)
+		return;
+
+	/* Unpinning the owned object is handled by generic shutdown */
+	if (!guc->g2g.owned)
+		xe_bo_put(guc->g2g.bo);
+
+	guc->g2g.bo = NULL;
+}
+
+static int guc_g2g_start(struct xe_guc *guc)
+{
+	struct xe_gt *far_gt, *gt = guc_to_gt(guc);
+	struct xe_device *xe = gt_to_xe(gt);
+	unsigned int i, j;
+	int t, err;
+	bool have_dev;
+
+	if (!guc->g2g.bo) {
+		int ret;
+
+		ret = guc_g2g_alloc(guc);
+		if (ret)
+			return ret;
+	}
+
+	/* GuC interface will need extending if more GT device types are ever created. */
+	xe_gt_assert(gt, (gt->info.type == XE_GT_TYPE_MAIN) || (gt->info.type == XE_GT_TYPE_MEDIA));
+
+	/* Channel numbering depends on whether there are multiple GTs per tile */
+	have_dev = xe->info.gt_count > xe->info.tile_count;
+
+	for_each_gt(far_gt, xe, i) {
+		u32 far_tile, far_dev;
+
+		if (far_gt->info.id == gt->info.id)
+			continue;
+
+		far_tile = gt_to_tile(far_gt)->id;
+		far_dev = G2G_DEV(far_gt);
+
+		for (t = 0; t < XE_G2G_TYPE_LIMIT; t++) {
+			err = guc_g2g_register(guc, far_gt, t, have_dev);
+			if (err) {
+				while (--t >= 0)
+					guc_g2g_deregister(guc, far_tile, far_dev, t);
+				goto err_deregister;
+			}
+		}
+	}
+
+	return 0;
+
+err_deregister:
+	for_each_gt(far_gt, xe, j) {
+		u32 tile, dev;
+
+		if (far_gt->info.id == gt->info.id)
+			continue;
+
+		if (j >= i)
+			break;
+
+		tile = gt_to_tile(far_gt)->id;
+		dev = G2G_DEV(far_gt);
+
+		for (t = 0; t < XE_G2G_TYPE_LIMIT; t++)
+			guc_g2g_deregister(guc, tile, dev, t);
+	}
+
+	return err;
+}
+
 static void guc_fini_hw(void *arg)
 {
 	struct xe_guc *guc = arg;
@@ -253,6 +548,8 @@ static void guc_fini_hw(void *arg)
 	fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
 	xe_uc_fini_hw(&guc_to_gt(guc)->uc);
 	xe_force_wake_put(gt_to_fw(gt), fw_ref);
+
+	guc_g2g_fini(guc);
 }
 
 /**
@@ -423,7 +720,16 @@ int xe_guc_init_post_hwconfig(struct xe_guc *guc)
 
 int xe_guc_post_load_init(struct xe_guc *guc)
 {
+	int ret;
+
 	xe_guc_ads_populate_post_load(&guc->ads);
+
+	if (xe_guc_g2g_wanted(guc_to_xe(guc))) {
+		ret = guc_g2g_start(guc);
+		if (ret)
+			return ret;
+	}
+
 	guc->submission_state.enabled = true;
 
 	return 0;
@@ -945,7 +1251,6 @@ int xe_guc_mmio_send_recv(struct xe_guc *guc, const u32 *request,
 
 	BUILD_BUG_ON(VF_SW_FLAG_COUNT != MED_VF_SW_FLAG_COUNT);
 
-	xe_assert(xe, !xe_guc_ct_enabled(&guc->ct));
 	xe_assert(xe, len);
 	xe_assert(xe, len <= VF_SW_FLAG_COUNT);
 	xe_assert(xe, len <= MED_VF_SW_FLAG_COUNT);
@@ -1099,10 +1404,21 @@ int xe_guc_self_cfg64(struct xe_guc *guc, u16 key, u64 val)
 	return guc_self_cfg(guc, key, 2, val);
 }
 
+static void xe_guc_sw_0_irq_handler(struct xe_guc *guc)
+{
+	struct xe_gt *gt = guc_to_gt(guc);
+
+	if (IS_SRIOV_VF(gt_to_xe(gt)))
+		xe_gt_sriov_vf_migrated_event_handler(gt);
+}
+
 void xe_guc_irq_handler(struct xe_guc *guc, const u16 iir)
 {
 	if (iir & GUC_INTR_GUC2HOST)
 		xe_guc_ct_irq_handler(&guc->ct);
+
+	if (iir & GUC_INTR_SW_INT_0)
+		xe_guc_sw_0_irq_handler(guc);
 }
 
 void xe_guc_sanitize(struct xe_guc *guc)
diff --git a/drivers/gpu/drm/xe/xe_guc_ads.c b/drivers/gpu/drm/xe/xe_guc_ads.c
index 4e746ae98888..943146e5b460 100644
--- a/drivers/gpu/drm/xe/xe_guc_ads.c
+++ b/drivers/gpu/drm/xe/xe_guc_ads.c
@@ -231,11 +231,6 @@ static size_t guc_ads_size(struct xe_guc_ads *ads)
 		guc_ads_private_data_size(ads);
 }
 
-static bool needs_wa_1607983814(struct xe_device *xe)
-{
-	return GRAPHICS_VERx100(xe) < 1250;
-}
-
 static size_t calculate_regset_size(struct xe_gt *gt)
 {
 	struct xe_reg_sr_entry *sr_entry;
@@ -250,7 +245,7 @@ static size_t calculate_regset_size(struct xe_gt *gt)
 
 	count += ADS_REGSET_EXTRA_MAX * XE_NUM_HW_ENGINES;
 
-	if (needs_wa_1607983814(gt_to_xe(gt)))
+	if (XE_WA(gt, 1607983814))
 		count += LNCFCMOCS_REG_COUNT;
 
 	return count * sizeof(struct guc_mmio_reg);
@@ -709,7 +704,6 @@ static unsigned int guc_mmio_regset_write(struct xe_guc_ads *ads,
 					  struct iosys_map *regset_map,
 					  struct xe_hw_engine *hwe)
 {
-	struct xe_device *xe = ads_to_xe(ads);
 	struct xe_hw_engine *hwe_rcs_reset_domain =
 		xe_gt_any_hw_engine_by_reset_domain(hwe->gt, XE_ENGINE_CLASS_RENDER);
 	struct xe_reg_sr_entry *entry;
@@ -740,8 +734,7 @@ static unsigned int guc_mmio_regset_write(struct xe_guc_ads *ads,
 		guc_mmio_regset_write_one(ads, regset_map, e->reg, count++);
 	}
 
-	/* Wa_1607983814 */
-	if (needs_wa_1607983814(xe) && hwe->class == XE_ENGINE_CLASS_RENDER) {
+	if (XE_WA(hwe->gt, 1607983814) && hwe->class == XE_ENGINE_CLASS_RENDER) {
 		for (i = 0; i < LNCFCMOCS_REG_COUNT; i++) {
 			guc_mmio_regset_write_one(ads, regset_map,
 						  XELP_LNCFCMOCS(i), count++);
diff --git a/drivers/gpu/drm/xe/xe_guc_capture.c b/drivers/gpu/drm/xe/xe_guc_capture.c
index d63912d28246..137571fae4ed 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture.c
+++ b/drivers/gpu/drm/xe/xe_guc_capture.c
@@ -1806,7 +1806,6 @@ void xe_engine_snapshot_print(struct xe_hw_engine_snapshot *snapshot, struct drm
 	if (!devcore_snapshot->matched_node)
 		return;
 
-	xe_gt_assert(gt, snapshot->source <= XE_ENGINE_CAPTURE_SOURCE_GUC);
 	xe_gt_assert(gt, snapshot->hwe);
 
 	capture_class = xe_engine_class_to_guc_capture_class(snapshot->hwe->class);
@@ -1815,7 +1814,8 @@ void xe_engine_snapshot_print(struct xe_hw_engine_snapshot *snapshot, struct drm
 		   snapshot->name ? snapshot->name : "",
 		   snapshot->logical_instance);
 	drm_printf(p, "\tCapture_source: %s\n",
-		   snapshot->source == XE_ENGINE_CAPTURE_SOURCE_GUC ? "GuC" : "Manual");
+		   devcore_snapshot->matched_node->source == XE_ENGINE_CAPTURE_SOURCE_GUC ?
+		   "GuC" : "Manual");
 	drm_printf(p, "\tCoverage: %s\n", grptype[devcore_snapshot->matched_node->is_partial]);
 	drm_printf(p, "\tForcewake: domain 0x%x, ref %d\n",
 		   snapshot->forcewake.domain, snapshot->forcewake.ref);
@@ -1840,29 +1840,24 @@ void xe_engine_snapshot_print(struct xe_hw_engine_snapshot *snapshot, struct drm
 }
 
 /**
- * xe_guc_capture_get_matching_and_lock - Matching GuC capture for the job.
- * @job: The job object.
+ * xe_guc_capture_get_matching_and_lock - Matching GuC capture for the queue.
+ * @q: The exec queue object
  *
- * Search within the capture outlist for the job, could be used for check if
- * GuC capture is ready for the job.
+ * Search within the capture outlist for the queue, could be used for check if
+ * GuC capture is ready for the queue.
  * If found, the locked boolean of the node will be flagged.
  *
  * Returns: found guc-capture node ptr else NULL
  */
 struct __guc_capture_parsed_output *
-xe_guc_capture_get_matching_and_lock(struct xe_sched_job *job)
+xe_guc_capture_get_matching_and_lock(struct xe_exec_queue *q)
 {
 	struct xe_hw_engine *hwe;
 	enum xe_hw_engine_id id;
-	struct xe_exec_queue *q;
 	struct xe_device *xe;
 	u16 guc_class = GUC_LAST_ENGINE_CLASS + 1;
 	struct xe_devcoredump_snapshot *ss;
 
-	if (!job)
-		return NULL;
-
-	q = job->q;
 	if (!q || !q->gt)
 		return NULL;
 
@@ -1874,7 +1869,7 @@ xe_guc_capture_get_matching_and_lock(struct xe_sched_job *job)
 	if (ss->matched_node && ss->matched_node->source == XE_ENGINE_CAPTURE_SOURCE_GUC)
 		return ss->matched_node;
 
-	/* Find hwe for the job */
+	/* Find hwe for the queue */
 	for_each_hw_engine(hwe, q->gt, id) {
 		if (hwe != q->hwe)
 			continue;
@@ -1906,17 +1901,16 @@ xe_guc_capture_get_matching_and_lock(struct xe_sched_job *job)
 }
 
 /**
- * xe_engine_snapshot_capture_for_job - Take snapshot of associated engine
- * @job: The job object
+ * xe_engine_snapshot_capture_for_queue - Take snapshot of associated engine
+ * @q: The exec queue object
  *
  * Take snapshot of associated HW Engine
  *
  * Returns: None.
  */
 void
-xe_engine_snapshot_capture_for_job(struct xe_sched_job *job)
+xe_engine_snapshot_capture_for_queue(struct xe_exec_queue *q)
 {
-	struct xe_exec_queue *q = job->q;
 	struct xe_device *xe = gt_to_xe(q->gt);
 	struct xe_devcoredump *coredump = &xe->devcoredump;
 	struct xe_hw_engine *hwe;
@@ -1934,11 +1928,12 @@ xe_engine_snapshot_capture_for_job(struct xe_sched_job *job)
 		}
 
 		if (!coredump->snapshot.hwe[id]) {
-			coredump->snapshot.hwe[id] = xe_hw_engine_snapshot_capture(hwe, job);
+			coredump->snapshot.hwe[id] =
+				xe_hw_engine_snapshot_capture(hwe, q);
 		} else {
 			struct __guc_capture_parsed_output *new;
 
-			new = xe_guc_capture_get_matching_and_lock(job);
+			new = xe_guc_capture_get_matching_and_lock(q);
 			if (new) {
 				struct xe_guc *guc =  &q->gt->uc.guc;
 
diff --git a/drivers/gpu/drm/xe/xe_guc_capture.h b/drivers/gpu/drm/xe/xe_guc_capture.h
index 97a795d13dd1..20a078dc4b85 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture.h
+++ b/drivers/gpu/drm/xe/xe_guc_capture.h
@@ -11,10 +11,10 @@
 #include "xe_guc.h"
 #include "xe_guc_fwif.h"
 
+struct xe_exec_queue;
 struct xe_guc;
 struct xe_hw_engine;
 struct xe_hw_engine_snapshot;
-struct xe_sched_job;
 
 static inline enum guc_capture_list_class_type xe_guc_class_to_capture_class(u16 class)
 {
@@ -50,10 +50,10 @@ size_t xe_guc_capture_ads_input_worst_size(struct xe_guc *guc);
 const struct __guc_mmio_reg_descr_group *
 xe_guc_capture_get_reg_desc_list(struct xe_gt *gt, u32 owner, u32 type,
 				 enum guc_capture_list_class_type capture_class, bool is_ext);
-struct __guc_capture_parsed_output *xe_guc_capture_get_matching_and_lock(struct xe_sched_job *job);
+struct __guc_capture_parsed_output *xe_guc_capture_get_matching_and_lock(struct xe_exec_queue *q);
 void xe_engine_manual_capture(struct xe_hw_engine *hwe, struct xe_hw_engine_snapshot *snapshot);
 void xe_engine_snapshot_print(struct xe_hw_engine_snapshot *snapshot, struct drm_printer *p);
-void xe_engine_snapshot_capture_for_job(struct xe_sched_job *job);
+void xe_engine_snapshot_capture_for_queue(struct xe_exec_queue *q);
 void xe_guc_capture_steered_list_init(struct xe_guc *guc);
 void xe_guc_capture_put_matched_nodes(struct xe_guc *guc);
 int xe_guc_capture_init(struct xe_guc *guc);
diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c
index 8aeb1789805c..7d33f3a11e61 100644
--- a/drivers/gpu/drm/xe/xe_guc_ct.c
+++ b/drivers/gpu/drm/xe/xe_guc_ct.c
@@ -54,6 +54,7 @@ enum {
 	CT_DEAD_PARSE_G2H_UNKNOWN,		/* 0x1000 */
 	CT_DEAD_PARSE_G2H_ORIGIN,		/* 0x2000 */
 	CT_DEAD_PARSE_G2H_TYPE,			/* 0x4000 */
+	CT_DEAD_CRASH,				/* 0x8000 */
 };
 
 static void ct_dead_worker_func(struct work_struct *w);
@@ -469,8 +470,10 @@ int xe_guc_ct_enable(struct xe_guc_ct *ct)
 	 * after any existing dead state has been dumped.
 	 */
 	spin_lock_irq(&ct->dead.lock);
-	if (ct->dead.reason)
+	if (ct->dead.reason) {
 		ct->dead.reason |= (1 << CT_DEAD_STATE_REARM);
+		queue_work(system_unbound_wq, &ct->dead.worker);
+	}
 	spin_unlock_irq(&ct->dead.lock);
 #endif
 
@@ -1017,7 +1020,6 @@ retry_same_fence:
 	}
 
 	ret = wait_event_timeout(ct->g2h_fence_wq, g2h_fence.done, HZ);
-
 	if (!ret) {
 		LNL_FLUSH_WORK(&ct->g2h_worker);
 		if (g2h_fence.done) {
@@ -1121,6 +1123,24 @@ static int parse_g2h_event(struct xe_guc_ct *ct, u32 *msg, u32 len)
 	return 0;
 }
 
+static int guc_crash_process_msg(struct xe_guc_ct *ct, u32 action)
+{
+	struct xe_gt *gt = ct_to_gt(ct);
+
+	if (action == XE_GUC_ACTION_NOTIFY_CRASH_DUMP_POSTED)
+		xe_gt_err(gt, "GuC Crash dump notification\n");
+	else if (action == XE_GUC_ACTION_NOTIFY_EXCEPTION)
+		xe_gt_err(gt, "GuC Exception notification\n");
+	else
+		xe_gt_err(gt, "Unknown GuC crash notification: 0x%04X\n", action);
+
+	CT_DEAD(ct, NULL, CRASH);
+
+	kick_reset(ct);
+
+	return 0;
+}
+
 static int parse_g2h_response(struct xe_guc_ct *ct, u32 *msg, u32 len)
 {
 	struct xe_gt *gt =  ct_to_gt(ct);
@@ -1295,13 +1315,17 @@ static int process_g2h_msg(struct xe_guc_ct *ct, u32 *msg, u32 len)
 	case GUC_ACTION_GUC2PF_ADVERSE_EVENT:
 		ret = xe_gt_sriov_pf_monitor_process_guc2pf(gt, hxg, hxg_len);
 		break;
+	case XE_GUC_ACTION_NOTIFY_CRASH_DUMP_POSTED:
+	case XE_GUC_ACTION_NOTIFY_EXCEPTION:
+		ret = guc_crash_process_msg(ct, action);
+		break;
 	default:
 		xe_gt_err(gt, "unexpected G2H action 0x%04x\n", action);
 	}
 
 	if (ret) {
-		xe_gt_err(gt, "G2H action 0x%04x failed (%pe)\n",
-			  action, ERR_PTR(ret));
+		xe_gt_err(gt, "G2H action %#04x failed (%pe) len %u msg %*ph\n",
+			  action, ERR_PTR(ret), hxg_len, (int)sizeof(u32) * hxg_len, hxg);
 		CT_DEAD(ct, NULL, PROCESS_FAILED);
 	}
 
diff --git a/drivers/gpu/drm/xe/xe_guc_fwif.h b/drivers/gpu/drm/xe/xe_guc_fwif.h
index 08ffe59f22fa..057153f89b30 100644
--- a/drivers/gpu/drm/xe/xe_guc_fwif.h
+++ b/drivers/gpu/drm/xe/xe_guc_fwif.h
@@ -17,6 +17,7 @@
 #define G2H_LEN_DW_TLB_INVALIDATE		3
 
 #define GUC_ID_MAX			65535
+#define GUC_ID_UNKNOWN			0xffffffff
 
 #define GUC_CONTEXT_DISABLE		0
 #define GUC_CONTEXT_ENABLE		1
diff --git a/drivers/gpu/drm/xe/xe_guc_klv_helpers.c b/drivers/gpu/drm/xe/xe_guc_klv_helpers.c
index 9d99fe266d97..146a6eda9e06 100644
--- a/drivers/gpu/drm/xe/xe_guc_klv_helpers.c
+++ b/drivers/gpu/drm/xe/xe_guc_klv_helpers.c
@@ -49,6 +49,8 @@ const char *xe_guc_klv_key_to_string(u16 key)
 		return "begin_db_id";
 	case GUC_KLV_VF_CFG_BEGIN_CONTEXT_ID_KEY:
 		return "begin_ctx_id";
+	case GUC_KLV_VF_CFG_SCHED_PRIORITY_KEY:
+		return "sched_priority";
 
 	/* VF CFG threshold keys */
 #define define_threshold_key_to_string_case(TAG, NAME, ...)	\
diff --git a/drivers/gpu/drm/xe/xe_guc_submit.c b/drivers/gpu/drm/xe/xe_guc_submit.c
index 6f4a9812b4f4..9c36329fe857 100644
--- a/drivers/gpu/drm/xe/xe_guc_submit.c
+++ b/drivers/gpu/drm/xe/xe_guc_submit.c
@@ -412,12 +412,11 @@ static const int xe_exec_queue_prio_to_guc[] = {
 static void init_policies(struct xe_guc *guc, struct xe_exec_queue *q)
 {
 	struct exec_queue_policy policy;
-	struct xe_device *xe = guc_to_xe(guc);
 	enum xe_exec_queue_priority prio = q->sched_props.priority;
 	u32 timeslice_us = q->sched_props.timeslice_us;
 	u32 preempt_timeout_us = q->sched_props.preempt_timeout_us;
 
-	xe_assert(xe, exec_queue_registered(q));
+	xe_gt_assert(guc_to_gt(guc), exec_queue_registered(q));
 
 	__guc_exec_queue_policy_start_klv(&policy, q->guc->id);
 	__guc_exec_queue_policy_add_priority(&policy, xe_exec_queue_prio_to_guc[prio]);
@@ -451,12 +450,11 @@ static void __register_mlrc_exec_queue(struct xe_guc *guc,
 				       struct guc_ctxt_registration_info *info)
 {
 #define MAX_MLRC_REG_SIZE      (13 + XE_HW_ENGINE_MAX_INSTANCE * 2)
-	struct xe_device *xe = guc_to_xe(guc);
 	u32 action[MAX_MLRC_REG_SIZE];
 	int len = 0;
 	int i;
 
-	xe_assert(xe, xe_exec_queue_is_parallel(q));
+	xe_gt_assert(guc_to_gt(guc), xe_exec_queue_is_parallel(q));
 
 	action[len++] = XE_GUC_ACTION_REGISTER_CONTEXT_MULTI_LRC;
 	action[len++] = info->flags;
@@ -479,7 +477,7 @@ static void __register_mlrc_exec_queue(struct xe_guc *guc,
 		action[len++] = upper_32_bits(xe_lrc_descriptor(lrc));
 	}
 
-	xe_assert(xe, len <= MAX_MLRC_REG_SIZE);
+	xe_gt_assert(guc_to_gt(guc), len <= MAX_MLRC_REG_SIZE);
 #undef MAX_MLRC_REG_SIZE
 
 	xe_guc_ct_send(&guc->ct, action, len, 0, 0);
@@ -513,7 +511,7 @@ static void register_exec_queue(struct xe_exec_queue *q)
 	struct xe_lrc *lrc = q->lrc[0];
 	struct guc_ctxt_registration_info info;
 
-	xe_assert(xe, !exec_queue_registered(q));
+	xe_gt_assert(guc_to_gt(guc), !exec_queue_registered(q));
 
 	memset(&info, 0, sizeof(info));
 	info.context_idx = q->guc->id;
@@ -603,7 +601,7 @@ static int wq_noop_append(struct xe_exec_queue *q)
 	if (wq_wait_for_space(q, wq_space_until_wrap(q)))
 		return -ENODEV;
 
-	xe_assert(xe, FIELD_FIT(WQ_LEN_MASK, len_dw));
+	xe_gt_assert(guc_to_gt(guc), FIELD_FIT(WQ_LEN_MASK, len_dw));
 
 	parallel_write(xe, map, wq[q->guc->wqi_tail / sizeof(u32)],
 		       FIELD_PREP(WQ_TYPE_MASK, WQ_TYPE_NOOP) |
@@ -643,13 +641,13 @@ static void wq_item_append(struct xe_exec_queue *q)
 		wqi[i++] = lrc->ring.tail / sizeof(u64);
 	}
 
-	xe_assert(xe, i == wqi_size / sizeof(u32));
+	xe_gt_assert(guc_to_gt(guc), i == wqi_size / sizeof(u32));
 
 	iosys_map_incr(&map, offsetof(struct guc_submit_parallel_scratch,
 				      wq[q->guc->wqi_tail / sizeof(u32)]));
 	xe_map_memcpy_to(xe, &map, 0, wqi, wqi_size);
 	q->guc->wqi_tail += wqi_size;
-	xe_assert(xe, q->guc->wqi_tail <= WQ_SIZE);
+	xe_gt_assert(guc_to_gt(guc), q->guc->wqi_tail <= WQ_SIZE);
 
 	xe_device_wmb(xe);
 
@@ -661,7 +659,6 @@ static void wq_item_append(struct xe_exec_queue *q)
 static void submit_exec_queue(struct xe_exec_queue *q)
 {
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_lrc *lrc = q->lrc[0];
 	u32 action[3];
 	u32 g2h_len = 0;
@@ -669,7 +666,7 @@ static void submit_exec_queue(struct xe_exec_queue *q)
 	int len = 0;
 	bool extra_submit = false;
 
-	xe_assert(xe, exec_queue_registered(q));
+	xe_gt_assert(guc_to_gt(guc), exec_queue_registered(q));
 
 	if (xe_exec_queue_is_parallel(q))
 		wq_item_append(q);
@@ -716,12 +713,11 @@ guc_exec_queue_run_job(struct drm_sched_job *drm_job)
 	struct xe_sched_job *job = to_xe_sched_job(drm_job);
 	struct xe_exec_queue *q = job->q;
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 	struct dma_fence *fence = NULL;
 	bool lr = xe_exec_queue_is_lr(q);
 
-	xe_assert(xe, !(exec_queue_destroyed(q) || exec_queue_pending_disable(q)) ||
-		  exec_queue_banned(q) || exec_queue_suspended(q));
+	xe_gt_assert(guc_to_gt(guc), !(exec_queue_destroyed(q) || exec_queue_pending_disable(q)) ||
+		     exec_queue_banned(q) || exec_queue_suspended(q));
 
 	trace_xe_sched_job_run(job);
 
@@ -823,7 +819,7 @@ static void xe_guc_exec_queue_trigger_cleanup(struct xe_exec_queue *q)
  */
 void xe_guc_submit_wedge(struct xe_guc *guc)
 {
-	struct xe_device *xe = guc_to_xe(guc);
+	struct xe_gt *gt = guc_to_gt(guc);
 	struct xe_exec_queue *q;
 	unsigned long index;
 	int err;
@@ -833,7 +829,8 @@ void xe_guc_submit_wedge(struct xe_guc *guc)
 	err = devm_add_action_or_reset(guc_to_xe(guc)->drm.dev,
 				       guc_submit_wedged_fini, guc);
 	if (err) {
-		drm_err(&xe->drm, "Failed to register xe_guc_submit clean-up on wedged.mode=2. Although device is wedged.\n");
+		xe_gt_err(gt, "Failed to register clean-up on wedged.mode=2; "
+			  "Although device is wedged.\n");
 		return;
 	}
 
@@ -865,11 +862,10 @@ static void xe_guc_exec_queue_lr_cleanup(struct work_struct *w)
 		container_of(w, struct xe_guc_exec_queue, lr_tdr);
 	struct xe_exec_queue *q = ge->q;
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_gpu_scheduler *sched = &ge->sched;
 	bool wedged;
 
-	xe_assert(xe, xe_exec_queue_is_lr(q));
+	xe_gt_assert(guc_to_gt(guc), xe_exec_queue_is_lr(q));
 	trace_xe_exec_queue_lr_cleanup(q);
 
 	wedged = guc_submit_hint_wedged(exec_queue_to_guc(q));
@@ -903,13 +899,19 @@ static void xe_guc_exec_queue_lr_cleanup(struct work_struct *w)
 					 !exec_queue_pending_disable(q) ||
 					 xe_guc_read_stopped(guc), HZ * 5);
 		if (!ret) {
-			drm_warn(&xe->drm, "Schedule disable failed to respond");
+			xe_gt_warn(q->gt, "Schedule disable failed to respond, guc_id=%d\n",
+				   q->guc->id);
+			xe_devcoredump(q, NULL, "Schedule disable failed to respond, guc_id=%d\n",
+				       q->guc->id);
 			xe_sched_submission_start(sched);
 			xe_gt_reset_async(q->gt);
 			return;
 		}
 	}
 
+	if (!exec_queue_killed(q) && !xe_lrc_ring_is_idle(q->lrc[0]))
+		xe_devcoredump(q, NULL, "LR job cleanup, guc_id=%d", q->guc->id);
+
 	xe_sched_submission_start(sched);
 }
 
@@ -1068,13 +1070,13 @@ guc_exec_queue_timedout_job(struct drm_sched_job *drm_job)
 	 * do manual capture first and decide later if we need to use it
 	 */
 	if (!exec_queue_killed(q) && !xe->devcoredump.captured &&
-	    !xe_guc_capture_get_matching_and_lock(job)) {
+	    !xe_guc_capture_get_matching_and_lock(q)) {
 		/* take force wake before engine register manual capture */
 		fw_ref = xe_force_wake_get(gt_to_fw(q->gt), XE_FORCEWAKE_ALL);
 		if (!xe_force_wake_ref_has_domain(fw_ref, XE_FORCEWAKE_ALL))
 			xe_gt_info(q->gt, "failed to get forcewake for coredump capture\n");
 
-		xe_engine_snapshot_capture_for_job(job);
+		xe_engine_snapshot_capture_for_queue(q);
 
 		xe_force_wake_put(gt_to_fw(q->gt), fw_ref);
 	}
@@ -1132,7 +1134,12 @@ guc_exec_queue_timedout_job(struct drm_sched_job *drm_job)
 		if (!ret || xe_guc_read_stopped(guc)) {
 trigger_reset:
 			if (!ret)
-				xe_gt_warn(guc_to_gt(guc), "Schedule disable failed to respond");
+				xe_gt_warn(guc_to_gt(guc),
+					   "Schedule disable failed to respond, guc_id=%d",
+					   q->guc->id);
+			xe_devcoredump(q, job,
+				       "Schedule disable failed to respond, guc_id=%d, ret=%d, guc_read=%d",
+				       q->guc->id, ret, xe_guc_read_stopped(guc));
 			set_exec_queue_extra_ref(q);
 			xe_exec_queue_get(q);	/* GT reset owns this */
 			set_exec_queue_banned(q);
@@ -1162,7 +1169,10 @@ trigger_reset:
 	trace_xe_sched_job_timedout(job);
 
 	if (!exec_queue_killed(q))
-		xe_devcoredump(job);
+		xe_devcoredump(q, job,
+			       "Timedout job - seqno=%u, lrc_seqno=%u, guc_id=%d, flags=0x%lx",
+			       xe_sched_job_seqno(job), xe_sched_job_lrc_seqno(job),
+			       q->guc->id, q->flags);
 
 	/*
 	 * Kernel jobs should never fail, nor should VM jobs if they do
@@ -1277,9 +1287,8 @@ static void __guc_exec_queue_process_msg_cleanup(struct xe_sched_msg *msg)
 {
 	struct xe_exec_queue *q = msg->private_data;
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 
-	xe_assert(xe, !(q->flags & EXEC_QUEUE_FLAG_PERMANENT));
+	xe_gt_assert(guc_to_gt(guc), !(q->flags & EXEC_QUEUE_FLAG_PERMANENT));
 	trace_xe_exec_queue_cleanup_entity(q);
 
 	if (exec_queue_registered(q))
@@ -1315,11 +1324,10 @@ static void __suspend_fence_signal(struct xe_exec_queue *q)
 static void suspend_fence_signal(struct xe_exec_queue *q)
 {
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 
-	xe_assert(xe, exec_queue_suspended(q) || exec_queue_killed(q) ||
-		  xe_guc_read_stopped(guc));
-	xe_assert(xe, q->guc->suspend_pending);
+	xe_gt_assert(guc_to_gt(guc), exec_queue_suspended(q) || exec_queue_killed(q) ||
+		     xe_guc_read_stopped(guc));
+	xe_gt_assert(guc_to_gt(guc), q->guc->suspend_pending);
 
 	__suspend_fence_signal(q);
 }
@@ -1415,12 +1423,11 @@ static int guc_exec_queue_init(struct xe_exec_queue *q)
 {
 	struct xe_gpu_scheduler *sched;
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_guc_exec_queue *ge;
 	long timeout;
 	int err, i;
 
-	xe_assert(xe, xe_device_uc_enabled(guc_to_xe(guc)));
+	xe_gt_assert(guc_to_gt(guc), xe_device_uc_enabled(guc_to_xe(guc)));
 
 	ge = kzalloc(sizeof(*ge), GFP_KERNEL);
 	if (!ge)
@@ -1633,9 +1640,8 @@ static void guc_exec_queue_resume(struct xe_exec_queue *q)
 	struct xe_gpu_scheduler *sched = &q->guc->sched;
 	struct xe_sched_msg *msg = q->guc->static_msgs + STATIC_MSG_RESUME;
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_device *xe = guc_to_xe(guc);
 
-	xe_assert(xe, !q->guc->suspend_pending);
+	xe_gt_assert(guc_to_gt(guc), !q->guc->suspend_pending);
 
 	xe_sched_msg_lock(sched);
 	guc_exec_queue_try_add_msg(q, msg, RESUME);
@@ -1708,7 +1714,7 @@ static void guc_exec_queue_stop(struct xe_guc *guc, struct xe_exec_queue *q)
 				ban = true;
 			}
 		} else if (xe_exec_queue_is_lr(q) &&
-			   (xe_lrc_ring_head(q->lrc[0]) != xe_lrc_ring_tail(q->lrc[0]))) {
+			   !xe_lrc_ring_is_idle(q->lrc[0])) {
 			ban = true;
 		}
 
@@ -1747,9 +1753,8 @@ void xe_guc_submit_stop(struct xe_guc *guc)
 {
 	struct xe_exec_queue *q;
 	unsigned long index;
-	struct xe_device *xe = guc_to_xe(guc);
 
-	xe_assert(xe, xe_guc_read_stopped(guc) == 1);
+	xe_gt_assert(guc_to_gt(guc), xe_guc_read_stopped(guc) == 1);
 
 	mutex_lock(&guc->submission_state.lock);
 
@@ -1791,9 +1796,8 @@ int xe_guc_submit_start(struct xe_guc *guc)
 {
 	struct xe_exec_queue *q;
 	unsigned long index;
-	struct xe_device *xe = guc_to_xe(guc);
 
-	xe_assert(xe, xe_guc_read_stopped(guc) == 1);
+	xe_gt_assert(guc_to_gt(guc), xe_guc_read_stopped(guc) == 1);
 
 	mutex_lock(&guc->submission_state.lock);
 	atomic_dec(&guc->submission_state.stopped);
@@ -1814,22 +1818,22 @@ int xe_guc_submit_start(struct xe_guc *guc)
 static struct xe_exec_queue *
 g2h_exec_queue_lookup(struct xe_guc *guc, u32 guc_id)
 {
-	struct xe_device *xe = guc_to_xe(guc);
+	struct xe_gt *gt = guc_to_gt(guc);
 	struct xe_exec_queue *q;
 
 	if (unlikely(guc_id >= GUC_ID_MAX)) {
-		drm_err(&xe->drm, "Invalid guc_id %u", guc_id);
+		xe_gt_err(gt, "Invalid guc_id %u\n", guc_id);
 		return NULL;
 	}
 
 	q = xa_load(&guc->submission_state.exec_queue_lookup, guc_id);
 	if (unlikely(!q)) {
-		drm_err(&xe->drm, "Not engine present for guc_id %u", guc_id);
+		xe_gt_err(gt, "Not engine present for guc_id %u\n", guc_id);
 		return NULL;
 	}
 
-	xe_assert(xe, guc_id >= q->guc->id);
-	xe_assert(xe, guc_id < (q->guc->id + q->width));
+	xe_gt_assert(guc_to_gt(guc), guc_id >= q->guc->id);
+	xe_gt_assert(guc_to_gt(guc), guc_id < (q->guc->id + q->width));
 
 	return q;
 }
@@ -1898,15 +1902,14 @@ static void handle_sched_done(struct xe_guc *guc, struct xe_exec_queue *q,
 
 int xe_guc_sched_done_handler(struct xe_guc *guc, u32 *msg, u32 len)
 {
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_exec_queue *q;
-	u32 guc_id = msg[0];
-	u32 runnable_state = msg[1];
+	u32 guc_id, runnable_state;
 
-	if (unlikely(len < 2)) {
-		drm_err(&xe->drm, "Invalid length %u", len);
+	if (unlikely(len < 2))
 		return -EPROTO;
-	}
+
+	guc_id = msg[0];
+	runnable_state = msg[1];
 
 	q = g2h_exec_queue_lookup(guc, guc_id);
 	if (unlikely(!q))
@@ -1940,14 +1943,13 @@ static void handle_deregister_done(struct xe_guc *guc, struct xe_exec_queue *q)
 
 int xe_guc_deregister_done_handler(struct xe_guc *guc, u32 *msg, u32 len)
 {
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_exec_queue *q;
-	u32 guc_id = msg[0];
+	u32 guc_id;
 
-	if (unlikely(len < 1)) {
-		drm_err(&xe->drm, "Invalid length %u", len);
+	if (unlikely(len < 1))
 		return -EPROTO;
-	}
+
+	guc_id = msg[0];
 
 	q = g2h_exec_queue_lookup(guc, guc_id);
 	if (unlikely(!q))
@@ -1969,14 +1971,13 @@ int xe_guc_deregister_done_handler(struct xe_guc *guc, u32 *msg, u32 len)
 int xe_guc_exec_queue_reset_handler(struct xe_guc *guc, u32 *msg, u32 len)
 {
 	struct xe_gt *gt = guc_to_gt(guc);
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_exec_queue *q;
-	u32 guc_id = msg[0];
+	u32 guc_id;
 
-	if (unlikely(len < 1)) {
-		drm_err(&xe->drm, "Invalid length %u", len);
+	if (unlikely(len < 1))
 		return -EPROTO;
-	}
+
+	guc_id = msg[0];
 
 	q = g2h_exec_queue_lookup(guc, guc_id);
 	if (unlikely(!q))
@@ -2016,10 +2017,8 @@ int xe_guc_error_capture_handler(struct xe_guc *guc, u32 *msg, u32 len)
 {
 	u32 status;
 
-	if (unlikely(len != XE_GUC_ACTION_STATE_CAPTURE_NOTIFICATION_DATA_LEN)) {
-		xe_gt_dbg(guc_to_gt(guc), "Invalid length %u", len);
+	if (unlikely(len != XE_GUC_ACTION_STATE_CAPTURE_NOTIFICATION_DATA_LEN))
 		return -EPROTO;
-	}
 
 	status = msg[0] & XE_GUC_STATE_CAPTURE_EVENT_STATUS_MASK;
 	if (status == XE_GUC_STATE_CAPTURE_EVENT_STATUS_NOSPACE)
@@ -2034,13 +2033,21 @@ int xe_guc_exec_queue_memory_cat_error_handler(struct xe_guc *guc, u32 *msg,
 					       u32 len)
 {
 	struct xe_gt *gt = guc_to_gt(guc);
-	struct xe_device *xe = guc_to_xe(guc);
 	struct xe_exec_queue *q;
-	u32 guc_id = msg[0];
+	u32 guc_id;
 
-	if (unlikely(len < 1)) {
-		drm_err(&xe->drm, "Invalid length %u", len);
+	if (unlikely(len < 1))
 		return -EPROTO;
+
+	guc_id = msg[0];
+
+	if (guc_id == GUC_ID_UNKNOWN) {
+		/*
+		 * GuC uses GUC_ID_UNKNOWN if it can not map the CAT fault to any PF/VF
+		 * context. In such case only PF will be notified about that fault.
+		 */
+		xe_gt_err_ratelimited(gt, "Memory CAT error reported by GuC!\n");
+		return 0;
 	}
 
 	q = g2h_exec_queue_lookup(guc, guc_id);
@@ -2062,24 +2069,22 @@ int xe_guc_exec_queue_memory_cat_error_handler(struct xe_guc *guc, u32 *msg,
 
 int xe_guc_exec_queue_reset_failure_handler(struct xe_guc *guc, u32 *msg, u32 len)
 {
-	struct xe_device *xe = guc_to_xe(guc);
+	struct xe_gt *gt = guc_to_gt(guc);
 	u8 guc_class, instance;
 	u32 reason;
 
-	if (unlikely(len != 3)) {
-		drm_err(&xe->drm, "Invalid length %u", len);
+	if (unlikely(len != 3))
 		return -EPROTO;
-	}
 
 	guc_class = msg[0];
 	instance = msg[1];
 	reason = msg[2];
 
 	/* Unexpected failure of a hardware feature, log an actual error */
-	drm_err(&xe->drm, "GuC engine reset request failed on %d:%d because 0x%08X",
-		guc_class, instance, reason);
+	xe_gt_err(gt, "GuC engine reset request failed on %d:%d because 0x%08X",
+		  guc_class, instance, reason);
 
-	xe_gt_reset_async(guc_to_gt(guc));
+	xe_gt_reset_async(gt);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/xe/xe_guc_types.h b/drivers/gpu/drm/xe/xe_guc_types.h
index fa75f57bf5da..83a41ebcdc91 100644
--- a/drivers/gpu/drm/xe/xe_guc_types.h
+++ b/drivers/gpu/drm/xe/xe_guc_types.h
@@ -64,6 +64,15 @@ struct xe_guc {
 	struct xe_guc_pc pc;
 	/** @dbm: GuC Doorbell Manager */
 	struct xe_guc_db_mgr dbm;
+
+	/** @g2g: GuC to GuC communication state */
+	struct {
+		/** @g2g.bo: Storage for GuC to GuC communication channels */
+		struct xe_bo *bo;
+		/** @g2g.owned: Is the BO owned by this GT or just mapped in */
+		bool owned;
+	} g2g;
+
 	/** @submission_state: GuC submission state */
 	struct {
 		/** @submission_state.idm: GuC context ID Manager */
@@ -79,6 +88,7 @@ struct xe_guc {
 		/** @submission_state.fini_wq: submit fini wait queue */
 		wait_queue_head_t fini_wq;
 	} submission_state;
+
 	/** @hwconfig: Hardware config state */
 	struct {
 		/** @hwconfig.bo: buffer object of the hardware config */
diff --git a/drivers/gpu/drm/xe/xe_heci_gsc.c b/drivers/gpu/drm/xe/xe_heci_gsc.c
index 65b2e147c4b9..d765bfd3636b 100644
--- a/drivers/gpu/drm/xe/xe_heci_gsc.c
+++ b/drivers/gpu/drm/xe/xe_heci_gsc.c
@@ -92,7 +92,7 @@ void xe_heci_gsc_fini(struct xe_device *xe)
 {
 	struct xe_heci_gsc *heci_gsc = &xe->heci_gsc;
 
-	if (!HAS_HECI_GSCFI(xe) && !HAS_HECI_CSCFI(xe))
+	if (!xe->info.has_heci_gscfi && !xe->info.has_heci_cscfi)
 		return;
 
 	if (heci_gsc->adev) {
@@ -177,7 +177,7 @@ void xe_heci_gsc_init(struct xe_device *xe)
 	const struct heci_gsc_def *def;
 	int ret;
 
-	if (!HAS_HECI_GSCFI(xe) && !HAS_HECI_CSCFI(xe))
+	if (!xe->info.has_heci_gscfi && !xe->info.has_heci_cscfi)
 		return;
 
 	heci_gsc->irq = -1;
@@ -222,7 +222,7 @@ void xe_heci_gsc_irq_handler(struct xe_device *xe, u32 iir)
 	if ((iir & GSC_IRQ_INTF(1)) == 0)
 		return;
 
-	if (!HAS_HECI_GSCFI(xe)) {
+	if (!xe->info.has_heci_gscfi) {
 		drm_warn_once(&xe->drm, "GSC irq: not supported");
 		return;
 	}
@@ -242,7 +242,7 @@ void xe_heci_csc_irq_handler(struct xe_device *xe, u32 iir)
 	if ((iir & CSC_IRQ_INTF(1)) == 0)
 		return;
 
-	if (!HAS_HECI_CSCFI(xe)) {
+	if (!xe->info.has_heci_cscfi) {
 		drm_warn_once(&xe->drm, "CSC irq: not supported");
 		return;
 	}
diff --git a/drivers/gpu/drm/xe/xe_hw_engine.c b/drivers/gpu/drm/xe/xe_hw_engine.c
index 1557acee3523..b19366744148 100644
--- a/drivers/gpu/drm/xe/xe_hw_engine.c
+++ b/drivers/gpu/drm/xe/xe_hw_engine.c
@@ -574,7 +574,6 @@ static int hw_engine_init(struct xe_gt *gt, struct xe_hw_engine *hwe,
 	xe_gt_assert(gt, gt->info.engine_mask & BIT(id));
 
 	xe_reg_sr_apply_mmio(&hwe->reg_sr, gt);
-	xe_reg_sr_apply_whitelist(hwe);
 
 	hwe->hwsp = xe_managed_bo_create_pin_map(xe, tile, SZ_4K,
 						 XE_BO_FLAG_VRAM_IF_DGFX(tile) |
@@ -829,7 +828,7 @@ void xe_hw_engine_handle_irq(struct xe_hw_engine *hwe, u16 intr_vec)
 /**
  * xe_hw_engine_snapshot_capture - Take a quick snapshot of the HW Engine.
  * @hwe: Xe HW Engine.
- * @job: The job object.
+ * @q: The exec queue object.
  *
  * This can be printed out in a later stage like during dev_coredump
  * analysis.
@@ -838,7 +837,7 @@ void xe_hw_engine_handle_irq(struct xe_hw_engine *hwe, u16 intr_vec)
  * caller, using `xe_hw_engine_snapshot_free`.
  */
 struct xe_hw_engine_snapshot *
-xe_hw_engine_snapshot_capture(struct xe_hw_engine *hwe, struct xe_sched_job *job)
+xe_hw_engine_snapshot_capture(struct xe_hw_engine *hwe, struct xe_exec_queue *q)
 {
 	struct xe_hw_engine_snapshot *snapshot;
 	struct __guc_capture_parsed_output *node;
@@ -864,15 +863,14 @@ xe_hw_engine_snapshot_capture(struct xe_hw_engine *hwe, struct xe_sched_job *job
 	if (IS_SRIOV_VF(gt_to_xe(hwe->gt)))
 		return snapshot;
 
-	if (job) {
+	if (q) {
 		/* If got guc capture, set source to GuC */
-		node = xe_guc_capture_get_matching_and_lock(job);
+		node = xe_guc_capture_get_matching_and_lock(q);
 		if (node) {
 			struct xe_device *xe = gt_to_xe(hwe->gt);
 			struct xe_devcoredump *coredump = &xe->devcoredump;
 
 			coredump->snapshot.matched_node = node;
-			snapshot->source = XE_ENGINE_CAPTURE_SOURCE_GUC;
 			xe_gt_dbg(hwe->gt, "Found and locked GuC-err-capture node");
 			return snapshot;
 		}
@@ -880,7 +878,6 @@ xe_hw_engine_snapshot_capture(struct xe_hw_engine *hwe, struct xe_sched_job *job
 
 	/* otherwise, do manual capture */
 	xe_engine_manual_capture(hwe, snapshot);
-	snapshot->source = XE_ENGINE_CAPTURE_SOURCE_MANUAL;
 	xe_gt_dbg(hwe->gt, "Proceeding with manual engine snapshot");
 
 	return snapshot;
diff --git a/drivers/gpu/drm/xe/xe_hw_engine.h b/drivers/gpu/drm/xe/xe_hw_engine.h
index da0a6922a26f..6b5f9fa2a594 100644
--- a/drivers/gpu/drm/xe/xe_hw_engine.h
+++ b/drivers/gpu/drm/xe/xe_hw_engine.h
@@ -11,7 +11,7 @@
 struct drm_printer;
 struct drm_xe_engine_class_instance;
 struct xe_device;
-struct xe_sched_job;
+struct xe_exec_queue;
 
 #ifdef CONFIG_DRM_XE_JOB_TIMEOUT_MIN
 #define XE_HW_ENGINE_JOB_TIMEOUT_MIN CONFIG_DRM_XE_JOB_TIMEOUT_MIN
@@ -56,7 +56,7 @@ void xe_hw_engine_enable_ring(struct xe_hw_engine *hwe);
 u32 xe_hw_engine_mask_per_class(struct xe_gt *gt,
 				enum xe_engine_class engine_class);
 struct xe_hw_engine_snapshot *
-xe_hw_engine_snapshot_capture(struct xe_hw_engine *hwe, struct xe_sched_job *job);
+xe_hw_engine_snapshot_capture(struct xe_hw_engine *hwe, struct xe_exec_queue *q);
 void xe_hw_engine_snapshot_free(struct xe_hw_engine_snapshot *snapshot);
 void xe_hw_engine_print(struct xe_hw_engine *hwe, struct drm_printer *p);
 void xe_hw_engine_setup_default_lrc_state(struct xe_hw_engine *hwe);
diff --git a/drivers/gpu/drm/xe/xe_hw_engine_types.h b/drivers/gpu/drm/xe/xe_hw_engine_types.h
index 719f27ef00a5..e14bee95e364 100644
--- a/drivers/gpu/drm/xe/xe_hw_engine_types.h
+++ b/drivers/gpu/drm/xe/xe_hw_engine_types.h
@@ -165,8 +165,6 @@ enum xe_hw_engine_snapshot_source_id {
 struct xe_hw_engine_snapshot {
 	/** @name: name of the hw engine */
 	char *name;
-	/** @source: Data source, either manual or GuC */
-	enum xe_hw_engine_snapshot_source_id source;
 	/** @hwe: hw engine */
 	struct xe_hw_engine *hwe;
 	/** @logical_instance: logical instance of this hw engine */
diff --git a/drivers/gpu/drm/xe/xe_irq.c b/drivers/gpu/drm/xe/xe_irq.c
index b7995ebd54ab..1c509e66694d 100644
--- a/drivers/gpu/drm/xe/xe_irq.c
+++ b/drivers/gpu/drm/xe/xe_irq.c
@@ -192,7 +192,7 @@ void xe_irq_enable_hwe(struct xe_gt *gt)
 		if (xe_hw_engine_mask_per_class(gt, XE_ENGINE_CLASS_OTHER)) {
 			gsc_mask = irqs | GSC_ER_COMPLETE;
 			heci_mask = GSC_IRQ_INTF(1);
-		} else if (HAS_HECI_GSCFI(xe)) {
+		} else if (xe->info.has_heci_gscfi) {
 			gsc_mask = GSC_IRQ_INTF(1);
 		}
 
@@ -325,7 +325,7 @@ static void gt_irq_handler(struct xe_tile *tile,
 
 			if (class == XE_ENGINE_CLASS_OTHER) {
 				/* HECI GSCFI interrupts come from outside of GT */
-				if (HAS_HECI_GSCFI(xe) && instance == OTHER_GSC_INSTANCE)
+				if (xe->info.has_heci_gscfi && instance == OTHER_GSC_INSTANCE)
 					xe_heci_gsc_irq_handler(xe, intr_vec);
 				else
 					gt_other_irq_handler(engine_gt, instance, intr_vec);
@@ -348,12 +348,8 @@ static irqreturn_t xelp_irq_handler(int irq, void *arg)
 	unsigned long intr_dw[2];
 	u32 identity[32];
 
-	spin_lock(&xe->irq.lock);
-	if (!xe->irq.enabled) {
-		spin_unlock(&xe->irq.lock);
+	if (!atomic_read(&xe->irq.enabled))
 		return IRQ_NONE;
-	}
-	spin_unlock(&xe->irq.lock);
 
 	master_ctl = xelp_intr_disable(xe);
 	if (!master_ctl) {
@@ -417,12 +413,8 @@ static irqreturn_t dg1_irq_handler(int irq, void *arg)
 
 	/* TODO: This really shouldn't be copied+pasted */
 
-	spin_lock(&xe->irq.lock);
-	if (!xe->irq.enabled) {
-		spin_unlock(&xe->irq.lock);
+	if (!atomic_read(&xe->irq.enabled))
 		return IRQ_NONE;
-	}
-	spin_unlock(&xe->irq.lock);
 
 	master_tile_ctl = dg1_intr_disable(xe);
 	if (!master_tile_ctl) {
@@ -459,7 +451,7 @@ static irqreturn_t dg1_irq_handler(int irq, void *arg)
 		 * the primary tile.
 		 */
 		if (id == 0) {
-			if (HAS_HECI_CSCFI(xe))
+			if (xe->info.has_heci_cscfi)
 				xe_heci_csc_irq_handler(xe, master_ctl);
 			xe_display_irq_handler(xe, master_ctl);
 			gu_misc_iir = gu_misc_irq_ack(xe, master_ctl);
@@ -508,7 +500,7 @@ static void gt_irq_reset(struct xe_tile *tile)
 
 	if ((tile->media_gt &&
 	     xe_hw_engine_mask_per_class(tile->media_gt, XE_ENGINE_CLASS_OTHER)) ||
-	    HAS_HECI_GSCFI(tile_to_xe(tile))) {
+	    tile_to_xe(tile)->info.has_heci_gscfi) {
 		xe_mmio_write32(mmio, GUNIT_GSC_INTR_ENABLE, 0);
 		xe_mmio_write32(mmio, GUNIT_GSC_INTR_MASK, ~0);
 		xe_mmio_write32(mmio, HECI2_RSVD_INTR_MASK, ~0);
@@ -644,12 +636,8 @@ static irqreturn_t vf_mem_irq_handler(int irq, void *arg)
 	struct xe_tile *tile;
 	unsigned int id;
 
-	spin_lock(&xe->irq.lock);
-	if (!xe->irq.enabled) {
-		spin_unlock(&xe->irq.lock);
+	if (!atomic_read(&xe->irq.enabled))
 		return IRQ_NONE;
-	}
-	spin_unlock(&xe->irq.lock);
 
 	for_each_tile(tile, xe, id)
 		xe_memirq_handler(&tile->memirq);
@@ -674,10 +662,9 @@ static void irq_uninstall(void *arg)
 	struct pci_dev *pdev = to_pci_dev(xe->drm.dev);
 	int irq;
 
-	if (!xe->irq.enabled)
+	if (!atomic_xchg(&xe->irq.enabled, 0))
 		return;
 
-	xe->irq.enabled = false;
 	xe_irq_reset(xe);
 
 	irq = pci_irq_vector(pdev, 0);
@@ -724,7 +711,7 @@ int xe_irq_install(struct xe_device *xe)
 		return err;
 	}
 
-	xe->irq.enabled = true;
+	atomic_set(&xe->irq.enabled, 1);
 
 	xe_irq_postinstall(xe);
 
@@ -744,9 +731,7 @@ void xe_irq_suspend(struct xe_device *xe)
 {
 	int irq = to_pci_dev(xe->drm.dev)->irq;
 
-	spin_lock_irq(&xe->irq.lock);
-	xe->irq.enabled = false; /* no new irqs */
-	spin_unlock_irq(&xe->irq.lock);
+	atomic_set(&xe->irq.enabled, 0); /* no new irqs */
 
 	synchronize_irq(irq); /* flush irqs */
 	xe_irq_reset(xe); /* turn irqs off */
@@ -762,7 +747,7 @@ void xe_irq_resume(struct xe_device *xe)
 	 * 1. no irq will arrive before the postinstall
 	 * 2. display is not yet resumed
 	 */
-	xe->irq.enabled = true;
+	atomic_set(&xe->irq.enabled, 1);
 	xe_irq_reset(xe);
 	xe_irq_postinstall(xe); /* turn irqs on */
 
diff --git a/drivers/gpu/drm/xe/xe_lrc.c b/drivers/gpu/drm/xe/xe_lrc.c
index 4f64c7f4e68d..22e58c6e2a35 100644
--- a/drivers/gpu/drm/xe/xe_lrc.c
+++ b/drivers/gpu/drm/xe/xe_lrc.c
@@ -25,6 +25,7 @@
 #include "xe_map.h"
 #include "xe_memirq.h"
 #include "xe_sriov.h"
+#include "xe_trace_lrc.h"
 #include "xe_vm.h"
 #include "xe_wa.h"
 
@@ -1060,6 +1061,14 @@ u32 xe_lrc_ring_tail(struct xe_lrc *lrc)
 		return xe_lrc_read_ctx_reg(lrc, CTX_RING_TAIL) & TAIL_ADDR;
 }
 
+static u32 xe_lrc_ring_start(struct xe_lrc *lrc)
+{
+	if (xe_lrc_has_indirect_ring_state(lrc))
+		return xe_lrc_read_indirect_ctx_reg(lrc, INDIRECT_CTX_RING_START);
+	else
+		return xe_lrc_read_ctx_reg(lrc, CTX_RING_START);
+}
+
 void xe_lrc_set_ring_head(struct xe_lrc *lrc, u32 head)
 {
 	if (xe_lrc_has_indirect_ring_state(lrc))
@@ -1635,10 +1644,12 @@ struct xe_lrc_snapshot *xe_lrc_snapshot_capture(struct xe_lrc *lrc)
 		xe_vm_get(lrc->bo->vm);
 
 	snapshot->context_desc = xe_lrc_ggtt_addr(lrc);
+	snapshot->ring_addr = __xe_lrc_ring_ggtt_addr(lrc);
 	snapshot->indirect_context_desc = xe_lrc_indirect_ring_ggtt_addr(lrc);
 	snapshot->head = xe_lrc_ring_head(lrc);
 	snapshot->tail.internal = lrc->ring.tail;
 	snapshot->tail.memory = xe_lrc_ring_tail(lrc);
+	snapshot->start = xe_lrc_ring_start(lrc);
 	snapshot->start_seqno = xe_lrc_start_seqno(lrc);
 	snapshot->seqno = xe_lrc_seqno(lrc);
 	snapshot->lrc_bo = xe_bo_get(lrc->bo);
@@ -1692,11 +1703,14 @@ void xe_lrc_snapshot_print(struct xe_lrc_snapshot *snapshot, struct drm_printer
 		return;
 
 	drm_printf(p, "\tHW Context Desc: 0x%08x\n", snapshot->context_desc);
+	drm_printf(p, "\tHW Ring address: 0x%08x\n",
+		   snapshot->ring_addr);
 	drm_printf(p, "\tHW Indirect Ring State: 0x%08x\n",
 		   snapshot->indirect_context_desc);
 	drm_printf(p, "\tLRC Head: (memory) %u\n", snapshot->head);
 	drm_printf(p, "\tLRC Tail: (internal) %u, (memory) %u\n",
 		   snapshot->tail.internal, snapshot->tail.memory);
+	drm_printf(p, "\tRing start: (memory) 0x%08x\n", snapshot->start);
 	drm_printf(p, "\tStart seqno: (memory) %d\n", snapshot->start_seqno);
 	drm_printf(p, "\tSeqno: (memory) %d\n", snapshot->seqno);
 	drm_printf(p, "\tTimestamp: 0x%08x\n", snapshot->ctx_timestamp);
@@ -1758,5 +1772,20 @@ u32 xe_lrc_update_timestamp(struct xe_lrc *lrc, u32 *old_ts)
 
 	lrc->ctx_timestamp = xe_lrc_ctx_timestamp(lrc);
 
+	trace_xe_lrc_update_timestamp(lrc, *old_ts);
+
 	return lrc->ctx_timestamp;
 }
+
+/**
+ * xe_lrc_ring_is_idle() - LRC is idle
+ * @lrc: Pointer to the lrc.
+ *
+ * Compare LRC ring head and tail to determine if idle.
+ *
+ * Return: True is ring is idle, False otherwise
+ */
+bool xe_lrc_ring_is_idle(struct xe_lrc *lrc)
+{
+	return xe_lrc_ring_head(lrc) == xe_lrc_ring_tail(lrc);
+}
diff --git a/drivers/gpu/drm/xe/xe_lrc.h b/drivers/gpu/drm/xe/xe_lrc.h
index 40d8f6906d3e..b459dcab8787 100644
--- a/drivers/gpu/drm/xe/xe_lrc.h
+++ b/drivers/gpu/drm/xe/xe_lrc.h
@@ -25,8 +25,10 @@ struct xe_lrc_snapshot {
 	unsigned long lrc_size, lrc_offset;
 
 	u32 context_desc;
+	u32 ring_addr;
 	u32 indirect_context_desc;
 	u32 head;
+	u32 start;
 	struct {
 		u32 internal;
 		u32 memory;
@@ -78,6 +80,8 @@ u32 xe_lrc_ring_head(struct xe_lrc *lrc);
 u32 xe_lrc_ring_space(struct xe_lrc *lrc);
 void xe_lrc_write_ring(struct xe_lrc *lrc, const void *data, size_t size);
 
+bool xe_lrc_ring_is_idle(struct xe_lrc *lrc);
+
 u32 xe_lrc_indirect_ring_ggtt_addr(struct xe_lrc *lrc);
 u32 xe_lrc_ggtt_addr(struct xe_lrc *lrc);
 u32 *xe_lrc_regs(struct xe_lrc *lrc);
diff --git a/drivers/gpu/drm/xe/xe_macros.h b/drivers/gpu/drm/xe/xe_macros.h
index daf56c846d03..8a77c2423555 100644
--- a/drivers/gpu/drm/xe/xe_macros.h
+++ b/drivers/gpu/drm/xe/xe_macros.h
@@ -10,9 +10,13 @@
 
 #define XE_WARN_ON WARN_ON
 
-#define XE_IOCTL_DBG(xe, cond) \
-	((cond) && (drm_dbg(&(xe)->drm, \
-			    "Ioctl argument check failed at %s:%d: %s", \
-			    __FILE__, __LINE__, #cond), 1))
+#define XE_IOCTL_DBG(xe, cond) ({					\
+	int cond__ = !!(cond);						\
+	if (cond__)							\
+		drm_dbg(&(xe)->drm,					\
+			"Ioctl argument check failed at %s:%d: %s",	\
+			__FILE__, __LINE__, #cond);			\
+	cond__;								\
+})
 
 #endif
diff --git a/drivers/gpu/drm/xe/xe_memirq.c b/drivers/gpu/drm/xe/xe_memirq.c
index f833da88150a..404fa2a456d5 100644
--- a/drivers/gpu/drm/xe/xe_memirq.c
+++ b/drivers/gpu/drm/xe/xe_memirq.c
@@ -155,13 +155,6 @@ static const char *guc_name(struct xe_guc *guc)
  *
  */
 
-static void __release_xe_bo(struct drm_device *drm, void *arg)
-{
-	struct xe_bo *bo = arg;
-
-	xe_bo_unpin_map_no_vm(bo);
-}
-
 static inline bool hw_reports_to_instance_zero(struct xe_memirq *memirq)
 {
 	/*
@@ -184,14 +177,12 @@ static int memirq_alloc_pages(struct xe_memirq *memirq)
 	BUILD_BUG_ON(!IS_ALIGNED(XE_MEMIRQ_SOURCE_OFFSET(0), SZ_64));
 	BUILD_BUG_ON(!IS_ALIGNED(XE_MEMIRQ_STATUS_OFFSET(0), SZ_4K));
 
-	/* XXX: convert to managed bo */
-	bo = xe_bo_create_pin_map(xe, tile, NULL, bo_size,
-				  ttm_bo_type_kernel,
-				  XE_BO_FLAG_SYSTEM |
-				  XE_BO_FLAG_GGTT |
-				  XE_BO_FLAG_GGTT_INVALIDATE |
-				  XE_BO_FLAG_NEEDS_UC |
-				  XE_BO_FLAG_NEEDS_CPU_ACCESS);
+	bo = xe_managed_bo_create_pin_map(xe, tile, bo_size,
+					  XE_BO_FLAG_SYSTEM |
+					  XE_BO_FLAG_GGTT |
+					  XE_BO_FLAG_GGTT_INVALIDATE |
+					  XE_BO_FLAG_NEEDS_UC |
+					  XE_BO_FLAG_NEEDS_CPU_ACCESS);
 	if (IS_ERR(bo)) {
 		err = PTR_ERR(bo);
 		goto out;
@@ -215,7 +206,7 @@ static int memirq_alloc_pages(struct xe_memirq *memirq)
 		     xe_bo_ggtt_addr(bo), bo_size, XE_MEMIRQ_SOURCE_OFFSET(0),
 		     XE_MEMIRQ_STATUS_OFFSET(0));
 
-	return drmm_add_action_or_reset(&xe->drm, __release_xe_bo, memirq->bo);
+	return 0;
 
 out:
 	memirq_err(memirq, "Failed to allocate memirq page (%pe)\n", ERR_PTR(err));
@@ -442,6 +433,9 @@ static void memirq_dispatch_guc(struct xe_memirq *memirq, struct iosys_map *stat
 
 	if (memirq_received(memirq, status, ilog2(GUC_INTR_GUC2HOST), name))
 		xe_guc_irq_handler(guc, GUC_INTR_GUC2HOST);
+
+	if (memirq_received(memirq, status, ilog2(GUC_INTR_SW_INT_0), name))
+		xe_guc_irq_handler(guc, GUC_INTR_SW_INT_0);
 }
 
 /**
diff --git a/drivers/gpu/drm/xe/xe_module.c b/drivers/gpu/drm/xe/xe_module.c
index bfc3deebdaa2..07b27114be9a 100644
--- a/drivers/gpu/drm/xe/xe_module.c
+++ b/drivers/gpu/drm/xe/xe_module.c
@@ -19,7 +19,7 @@
 
 struct xe_modparam xe_modparam = {
 	.probe_display = true,
-	.guc_log_level = 5,
+	.guc_log_level = 3,
 	.force_probe = CONFIG_DRM_XE_FORCE_PROBE,
 	.wedged_mode = 1,
 	/* the rest are 0 by default */
diff --git a/drivers/gpu/drm/xe/xe_oa.c b/drivers/gpu/drm/xe/xe_oa.c
index 8dd55798ab31..ec88b18e9baa 100644
--- a/drivers/gpu/drm/xe/xe_oa.c
+++ b/drivers/gpu/drm/xe/xe_oa.c
@@ -96,6 +96,7 @@ struct xe_oa_open_param {
 	struct drm_xe_sync __user *syncs_user;
 	int num_syncs;
 	struct xe_sync_entry *syncs;
+	size_t oa_buffer_size;
 };
 
 struct xe_oa_config_bo {
@@ -403,11 +404,19 @@ static int xe_oa_append_reports(struct xe_oa_stream *stream, char __user *buf,
 
 static void xe_oa_init_oa_buffer(struct xe_oa_stream *stream)
 {
-	struct xe_mmio *mmio = &stream->gt->mmio;
 	u32 gtt_offset = xe_bo_ggtt_addr(stream->oa_buffer.bo);
-	u32 oa_buf = gtt_offset | OABUFFER_SIZE_16M | OAG_OABUFFER_MEMORY_SELECT;
+	int size_exponent = __ffs(stream->oa_buffer.bo->size);
+	u32 oa_buf = gtt_offset | OAG_OABUFFER_MEMORY_SELECT;
+	struct xe_mmio *mmio = &stream->gt->mmio;
 	unsigned long flags;
 
+	/*
+	 * If oa buffer size is more than 16MB (exponent greater than 24), the
+	 * oa buffer size field is multiplied by 8 in xe_oa_enable_metric_set.
+	 */
+	oa_buf |= REG_FIELD_PREP(OABUFFER_SIZE_MASK,
+		size_exponent > 24 ? size_exponent - 20 : size_exponent - 17);
+
 	spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags);
 
 	xe_mmio_write32(mmio, __oa_regs(stream)->oa_status, 0);
@@ -901,15 +910,12 @@ static void xe_oa_stream_destroy(struct xe_oa_stream *stream)
 	xe_file_put(stream->xef);
 }
 
-static int xe_oa_alloc_oa_buffer(struct xe_oa_stream *stream)
+static int xe_oa_alloc_oa_buffer(struct xe_oa_stream *stream, size_t size)
 {
 	struct xe_bo *bo;
 
-	BUILD_BUG_ON_NOT_POWER_OF_2(XE_OA_BUFFER_SIZE);
-	BUILD_BUG_ON(XE_OA_BUFFER_SIZE < SZ_128K || XE_OA_BUFFER_SIZE > SZ_16M);
-
 	bo = xe_bo_create_pin_map(stream->oa->xe, stream->gt->tile, NULL,
-				  XE_OA_BUFFER_SIZE, ttm_bo_type_kernel,
+				  size, ttm_bo_type_kernel,
 				  XE_BO_FLAG_SYSTEM | XE_BO_FLAG_GGTT);
 	if (IS_ERR(bo))
 		return PTR_ERR(bo);
@@ -1087,6 +1093,13 @@ static u32 oag_report_ctx_switches(const struct xe_oa_stream *stream)
 			     0 : OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS);
 }
 
+static u32 oag_buf_size_select(const struct xe_oa_stream *stream)
+{
+	return _MASKED_FIELD(OAG_OA_DEBUG_BUF_SIZE_SELECT,
+			     stream->oa_buffer.bo->size > SZ_16M ?
+			     OAG_OA_DEBUG_BUF_SIZE_SELECT : 0);
+}
+
 static int xe_oa_enable_metric_set(struct xe_oa_stream *stream)
 {
 	struct xe_mmio *mmio = &stream->gt->mmio;
@@ -1119,6 +1132,7 @@ static int xe_oa_enable_metric_set(struct xe_oa_stream *stream)
 	xe_mmio_write32(mmio, __oa_regs(stream)->oa_debug,
 			_MASKED_BIT_ENABLE(oa_debug) |
 			oag_report_ctx_switches(stream) |
+			oag_buf_size_select(stream) |
 			oag_configure_mmio_trigger(stream, true));
 
 	xe_mmio_write32(mmio, __oa_regs(stream)->oa_ctx_ctrl, stream->periodic ?
@@ -1260,6 +1274,17 @@ static int xe_oa_set_prop_syncs_user(struct xe_oa *oa, u64 value,
 	return 0;
 }
 
+static int xe_oa_set_prop_oa_buffer_size(struct xe_oa *oa, u64 value,
+					 struct xe_oa_open_param *param)
+{
+	if (!is_power_of_2(value) || value < SZ_128K || value > SZ_128M) {
+		drm_dbg(&oa->xe->drm, "OA buffer size invalid %llu\n", value);
+		return -EINVAL;
+	}
+	param->oa_buffer_size = value;
+	return 0;
+}
+
 static int xe_oa_set_prop_ret_inval(struct xe_oa *oa, u64 value,
 				    struct xe_oa_open_param *param)
 {
@@ -1280,6 +1305,7 @@ static const xe_oa_set_property_fn xe_oa_set_property_funcs_open[] = {
 	[DRM_XE_OA_PROPERTY_NO_PREEMPT] = xe_oa_set_no_preempt,
 	[DRM_XE_OA_PROPERTY_NUM_SYNCS] = xe_oa_set_prop_num_syncs,
 	[DRM_XE_OA_PROPERTY_SYNCS] = xe_oa_set_prop_syncs_user,
+	[DRM_XE_OA_PROPERTY_OA_BUFFER_SIZE] = xe_oa_set_prop_oa_buffer_size,
 };
 
 static const xe_oa_set_property_fn xe_oa_set_property_funcs_config[] = {
@@ -1294,6 +1320,7 @@ static const xe_oa_set_property_fn xe_oa_set_property_funcs_config[] = {
 	[DRM_XE_OA_PROPERTY_NO_PREEMPT] = xe_oa_set_prop_ret_inval,
 	[DRM_XE_OA_PROPERTY_NUM_SYNCS] = xe_oa_set_prop_num_syncs,
 	[DRM_XE_OA_PROPERTY_SYNCS] = xe_oa_set_prop_syncs_user,
+	[DRM_XE_OA_PROPERTY_OA_BUFFER_SIZE] = xe_oa_set_prop_ret_inval,
 };
 
 static int xe_oa_user_ext_set_property(struct xe_oa *oa, enum xe_oa_user_extn_from from,
@@ -1553,7 +1580,7 @@ static long xe_oa_status_locked(struct xe_oa_stream *stream, unsigned long arg)
 
 static long xe_oa_info_locked(struct xe_oa_stream *stream, unsigned long arg)
 {
-	struct drm_xe_oa_stream_info info = { .oa_buf_size = XE_OA_BUFFER_SIZE, };
+	struct drm_xe_oa_stream_info info = { .oa_buf_size = stream->oa_buffer.bo->size, };
 	void __user *uaddr = (void __user *)arg;
 
 	if (copy_to_user(uaddr, &info, sizeof(info)))
@@ -1639,7 +1666,7 @@ static int xe_oa_mmap(struct file *file, struct vm_area_struct *vma)
 	}
 
 	/* Can mmap the entire OA buffer or nothing (no partial OA buffer mmaps) */
-	if (vma->vm_end - vma->vm_start != XE_OA_BUFFER_SIZE) {
+	if (vma->vm_end - vma->vm_start != stream->oa_buffer.bo->size) {
 		drm_dbg(&stream->oa->xe->drm, "Wrong mmap size, must be OA buffer size\n");
 		return -EINVAL;
 	}
@@ -1783,9 +1810,10 @@ static int xe_oa_stream_init(struct xe_oa_stream *stream,
 	if (GRAPHICS_VER(stream->oa->xe) >= 20 &&
 	    stream->hwe->oa_unit->type == DRM_XE_OA_UNIT_TYPE_OAG && stream->sample)
 		stream->oa_buffer.circ_size =
-			XE_OA_BUFFER_SIZE - XE_OA_BUFFER_SIZE % stream->oa_buffer.format->size;
+			param->oa_buffer_size -
+			param->oa_buffer_size % stream->oa_buffer.format->size;
 	else
-		stream->oa_buffer.circ_size = XE_OA_BUFFER_SIZE;
+		stream->oa_buffer.circ_size = param->oa_buffer_size;
 
 	if (stream->exec_q && engine_supports_mi_query(stream->hwe)) {
 		/* If we don't find the context offset, just return error */
@@ -1828,7 +1856,7 @@ static int xe_oa_stream_init(struct xe_oa_stream *stream,
 		goto err_fw_put;
 	}
 
-	ret = xe_oa_alloc_oa_buffer(stream);
+	ret = xe_oa_alloc_oa_buffer(stream, param->oa_buffer_size);
 	if (ret)
 		goto err_fw_put;
 
@@ -2125,6 +2153,9 @@ int xe_oa_stream_open_ioctl(struct drm_device *dev, u64 data, struct drm_file *f
 		drm_dbg(&oa->xe->drm, "Using periodic sampling freq %lld Hz\n", oa_freq_hz);
 	}
 
+	if (!param.oa_buffer_size)
+		param.oa_buffer_size = DEFAULT_XE_OA_BUFFER_SIZE;
+
 	ret = xe_oa_parse_syncs(oa, &param);
 	if (ret)
 		goto err_exec_q;
diff --git a/drivers/gpu/drm/xe/xe_oa_types.h b/drivers/gpu/drm/xe/xe_oa_types.h
index fea9d981e414..df7793915628 100644
--- a/drivers/gpu/drm/xe/xe_oa_types.h
+++ b/drivers/gpu/drm/xe/xe_oa_types.h
@@ -15,7 +15,7 @@
 #include "regs/xe_reg_defs.h"
 #include "xe_hw_engine_types.h"
 
-#define XE_OA_BUFFER_SIZE SZ_16M
+#define DEFAULT_XE_OA_BUFFER_SIZE SZ_16M
 
 enum xe_oa_report_header {
 	HDR_32_BIT = 0,
diff --git a/drivers/gpu/drm/xe/xe_pm.c b/drivers/gpu/drm/xe/xe_pm.c
index 40f7c844ed44..80699dbeb2e9 100644
--- a/drivers/gpu/drm/xe/xe_pm.c
+++ b/drivers/gpu/drm/xe/xe_pm.c
@@ -738,9 +738,6 @@ void xe_pm_d3cold_allowed_toggle(struct xe_device *xe)
 		xe->d3cold.allowed = false;
 
 	mutex_unlock(&xe->d3cold.lock);
-
-	drm_dbg(&xe->drm,
-		"d3cold: allowed=%s\n", str_yes_no(xe->d3cold.allowed));
 }
 
 /**
diff --git a/drivers/gpu/drm/xe/xe_pt.c b/drivers/gpu/drm/xe/xe_pt.c
index f27f579f4d85..65c3c1688710 100644
--- a/drivers/gpu/drm/xe/xe_pt.c
+++ b/drivers/gpu/drm/xe/xe_pt.c
@@ -136,6 +136,7 @@ err_kfree:
 	xe_pt_free(pt);
 	return ERR_PTR(err);
 }
+ALLOW_ERROR_INJECTION(xe_pt_create, ERRNO);
 
 /**
  * xe_pt_populate_empty() - Populate a page-table bo with scratch- or zero
@@ -1333,8 +1334,7 @@ static void invalidation_fence_cb(struct dma_fence *fence,
 		queue_work(system_wq, &ifence->work);
 	} else {
 		ifence->base.base.error = ifence->fence->error;
-		dma_fence_signal(&ifence->base.base);
-		dma_fence_put(&ifence->base.base);
+		xe_gt_tlb_invalidation_fence_signal(&ifence->base);
 	}
 	dma_fence_put(ifence->fence);
 }
@@ -1851,6 +1851,7 @@ int xe_pt_update_ops_prepare(struct xe_tile *tile, struct xe_vma_ops *vops)
 
 	return 0;
 }
+ALLOW_ERROR_INJECTION(xe_pt_update_ops_prepare, ERRNO);
 
 static void bind_op_commit(struct xe_vm *vm, struct xe_tile *tile,
 			   struct xe_vm_pgtable_update_ops *pt_update_ops,
@@ -2131,6 +2132,7 @@ kill_vm_tile1:
 
 	return ERR_PTR(err);
 }
+ALLOW_ERROR_INJECTION(xe_pt_update_ops_run, ERRNO);
 
 /**
  * xe_pt_update_ops_fini() - Finish PT update operations
diff --git a/drivers/gpu/drm/xe/xe_query.c b/drivers/gpu/drm/xe/xe_query.c
index 170ae72d1a7b..d2a816f71bf2 100644
--- a/drivers/gpu/drm/xe/xe_query.c
+++ b/drivers/gpu/drm/xe/xe_query.c
@@ -23,6 +23,7 @@
 #include "xe_guc_hwconfig.h"
 #include "xe_macros.h"
 #include "xe_mmio.h"
+#include "xe_oa.h"
 #include "xe_ttm_vram_mgr.h"
 #include "xe_wa.h"
 
@@ -670,7 +671,8 @@ static int query_oa_units(struct xe_device *xe,
 			du->oa_unit_id = u->oa_unit_id;
 			du->oa_unit_type = u->type;
 			du->oa_timestamp_freq = xe_oa_timestamp_frequency(gt);
-			du->capabilities = DRM_XE_OA_CAPS_BASE | DRM_XE_OA_CAPS_SYNCS;
+			du->capabilities = DRM_XE_OA_CAPS_BASE | DRM_XE_OA_CAPS_SYNCS |
+					   DRM_XE_OA_CAPS_OA_BUFFER_SIZE;
 
 			j = 0;
 			for_each_hw_engine(hwe, gt, hwe_id) {
diff --git a/drivers/gpu/drm/xe/xe_reg_sr.c b/drivers/gpu/drm/xe/xe_reg_sr.c
index e1a0e27cda14..9475e3f74958 100644
--- a/drivers/gpu/drm/xe/xe_reg_sr.c
+++ b/drivers/gpu/drm/xe/xe_reg_sr.c
@@ -24,49 +24,29 @@
 #include "xe_hw_engine_types.h"
 #include "xe_macros.h"
 #include "xe_mmio.h"
-#include "xe_reg_whitelist.h"
 #include "xe_rtp_types.h"
 
-#define XE_REG_SR_GROW_STEP_DEFAULT	16
-
 static void reg_sr_fini(struct drm_device *drm, void *arg)
 {
 	struct xe_reg_sr *sr = arg;
+	struct xe_reg_sr_entry *entry;
+	unsigned long reg;
+
+	xa_for_each(&sr->xa, reg, entry)
+		kfree(entry);
 
 	xa_destroy(&sr->xa);
-	kfree(sr->pool.arr);
-	memset(&sr->pool, 0, sizeof(sr->pool));
 }
 
 int xe_reg_sr_init(struct xe_reg_sr *sr, const char *name, struct xe_device *xe)
 {
 	xa_init(&sr->xa);
-	memset(&sr->pool, 0, sizeof(sr->pool));
-	sr->pool.grow_step = XE_REG_SR_GROW_STEP_DEFAULT;
 	sr->name = name;
 
 	return drmm_add_action_or_reset(&xe->drm, reg_sr_fini, sr);
 }
 EXPORT_SYMBOL_IF_KUNIT(xe_reg_sr_init);
 
-static struct xe_reg_sr_entry *alloc_entry(struct xe_reg_sr *sr)
-{
-	if (sr->pool.used == sr->pool.allocated) {
-		struct xe_reg_sr_entry *arr;
-
-		arr = krealloc_array(sr->pool.arr,
-				     ALIGN(sr->pool.allocated + 1, sr->pool.grow_step),
-				     sizeof(*arr), GFP_KERNEL);
-		if (!arr)
-			return NULL;
-
-		sr->pool.arr = arr;
-		sr->pool.allocated += sr->pool.grow_step;
-	}
-
-	return &sr->pool.arr[sr->pool.used++];
-}
-
 static bool compatible_entries(const struct xe_reg_sr_entry *e1,
 			       const struct xe_reg_sr_entry *e2)
 {
@@ -112,7 +92,7 @@ int xe_reg_sr_add(struct xe_reg_sr *sr,
 		return 0;
 	}
 
-	pentry = alloc_entry(sr);
+	pentry = kmalloc(sizeof(*pentry), GFP_KERNEL);
 	if (!pentry) {
 		ret = -ENOMEM;
 		goto fail;
@@ -211,58 +191,6 @@ err_force_wake:
 	xe_gt_err(gt, "Failed to apply, err=-ETIMEDOUT\n");
 }
 
-void xe_reg_sr_apply_whitelist(struct xe_hw_engine *hwe)
-{
-	struct xe_reg_sr *sr = &hwe->reg_whitelist;
-	struct xe_gt *gt = hwe->gt;
-	struct xe_device *xe = gt_to_xe(gt);
-	struct xe_reg_sr_entry *entry;
-	struct drm_printer p;
-	u32 mmio_base = hwe->mmio_base;
-	unsigned long reg;
-	unsigned int slot = 0;
-	unsigned int fw_ref;
-
-	if (xa_empty(&sr->xa))
-		return;
-
-	drm_dbg(&xe->drm, "Whitelisting %s registers\n", sr->name);
-
-	fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
-	if (!xe_force_wake_ref_has_domain(fw_ref, XE_FORCEWAKE_ALL))
-		goto err_force_wake;
-
-	p = drm_dbg_printer(&xe->drm, DRM_UT_DRIVER, NULL);
-	xa_for_each(&sr->xa, reg, entry) {
-		if (slot == RING_MAX_NONPRIV_SLOTS) {
-			xe_gt_err(gt,
-				  "hwe %s: maximum register whitelist slots (%d) reached, refusing to add more\n",
-				  hwe->name, RING_MAX_NONPRIV_SLOTS);
-			break;
-		}
-
-		xe_reg_whitelist_print_entry(&p, 0, reg, entry);
-		xe_mmio_write32(&gt->mmio, RING_FORCE_TO_NONPRIV(mmio_base, slot),
-				reg | entry->set_bits);
-		slot++;
-	}
-
-	/* And clear the rest just in case of garbage */
-	for (; slot < RING_MAX_NONPRIV_SLOTS; slot++) {
-		u32 addr = RING_NOPID(mmio_base).addr;
-
-		xe_mmio_write32(&gt->mmio, RING_FORCE_TO_NONPRIV(mmio_base, slot), addr);
-	}
-
-	xe_force_wake_put(gt_to_fw(gt), fw_ref);
-
-	return;
-
-err_force_wake:
-	xe_force_wake_put(gt_to_fw(gt), fw_ref);
-	drm_err(&xe->drm, "Failed to apply, err=-ETIMEDOUT\n");
-}
-
 /**
  * xe_reg_sr_dump - print all save/restore entries
  * @sr: Save/restore entries
diff --git a/drivers/gpu/drm/xe/xe_reg_sr_types.h b/drivers/gpu/drm/xe/xe_reg_sr_types.h
index ad48a52b824a..ebe11f237fa2 100644
--- a/drivers/gpu/drm/xe/xe_reg_sr_types.h
+++ b/drivers/gpu/drm/xe/xe_reg_sr_types.h
@@ -20,12 +20,6 @@ struct xe_reg_sr_entry {
 };
 
 struct xe_reg_sr {
-	struct {
-		struct xe_reg_sr_entry *arr;
-		unsigned int used;
-		unsigned int allocated;
-		unsigned int grow_step;
-	} pool;
 	struct xarray xa;
 	const char *name;
 
diff --git a/drivers/gpu/drm/xe/xe_reg_whitelist.c b/drivers/gpu/drm/xe/xe_reg_whitelist.c
index 3996934974fa..edab5d4e3ba5 100644
--- a/drivers/gpu/drm/xe/xe_reg_whitelist.c
+++ b/drivers/gpu/drm/xe/xe_reg_whitelist.c
@@ -10,7 +10,9 @@
 #include "regs/xe_oa_regs.h"
 #include "regs/xe_regs.h"
 #include "xe_gt_types.h"
+#include "xe_gt_printk.h"
 #include "xe_platform_types.h"
+#include "xe_reg_sr.h"
 #include "xe_rtp.h"
 #include "xe_step.h"
 
@@ -89,6 +91,40 @@ static const struct xe_rtp_entry_sr register_whitelist[] = {
 	{}
 };
 
+static void whitelist_apply_to_hwe(struct xe_hw_engine *hwe)
+{
+	struct xe_reg_sr *sr = &hwe->reg_whitelist;
+	struct xe_reg_sr_entry *entry;
+	struct drm_printer p;
+	unsigned long reg;
+	unsigned int slot;
+
+	xe_gt_dbg(hwe->gt, "Add %s whitelist to engine\n", sr->name);
+	p = xe_gt_dbg_printer(hwe->gt);
+
+	slot = 0;
+	xa_for_each(&sr->xa, reg, entry) {
+		struct xe_reg_sr_entry hwe_entry = {
+			.reg = RING_FORCE_TO_NONPRIV(hwe->mmio_base, slot),
+			.set_bits = entry->reg.addr | entry->set_bits,
+			.clr_bits = ~0u,
+			.read_mask = entry->read_mask,
+		};
+
+		if (slot == RING_MAX_NONPRIV_SLOTS) {
+			xe_gt_err(hwe->gt,
+				  "hwe %s: maximum register whitelist slots (%d) reached, refusing to add more\n",
+				  hwe->name, RING_MAX_NONPRIV_SLOTS);
+			break;
+		}
+
+		xe_reg_whitelist_print_entry(&p, 0, reg, entry);
+		xe_reg_sr_add(&hwe->reg_sr, &hwe_entry, hwe->gt);
+
+		slot++;
+	}
+}
+
 /**
  * xe_reg_whitelist_process_engine - process table of registers to whitelist
  * @hwe: engine instance to process whitelist for
@@ -102,6 +138,7 @@ void xe_reg_whitelist_process_engine(struct xe_hw_engine *hwe)
 	struct xe_rtp_process_ctx ctx = XE_RTP_PROCESS_CTX_INITIALIZER(hwe);
 
 	xe_rtp_process_to_sr(&ctx, register_whitelist, &hwe->reg_whitelist);
+	whitelist_apply_to_hwe(hwe);
 }
 
 /**
diff --git a/drivers/gpu/drm/xe/xe_sriov.c b/drivers/gpu/drm/xe/xe_sriov.c
index ef10782af656..04e2f539ccd9 100644
--- a/drivers/gpu/drm/xe/xe_sriov.c
+++ b/drivers/gpu/drm/xe/xe_sriov.c
@@ -14,6 +14,7 @@
 #include "xe_mmio.h"
 #include "xe_sriov.h"
 #include "xe_sriov_pf.h"
+#include "xe_sriov_vf.h"
 
 /**
  * xe_sriov_mode_to_string - Convert enum value to string.
@@ -114,6 +115,9 @@ int xe_sriov_init(struct xe_device *xe)
 			return err;
 	}
 
+	if (IS_SRIOV_VF(xe))
+		xe_sriov_vf_init_early(xe);
+
 	xe_assert(xe, !xe->sriov.wq);
 	xe->sriov.wq = alloc_workqueue("xe-sriov-wq", 0, 0);
 	if (!xe->sriov.wq)
diff --git a/drivers/gpu/drm/xe/xe_sriov_pf_helpers.h b/drivers/gpu/drm/xe/xe_sriov_pf_helpers.h
index 7d156ba82479..dd1df950b021 100644
--- a/drivers/gpu/drm/xe/xe_sriov_pf_helpers.h
+++ b/drivers/gpu/drm/xe/xe_sriov_pf_helpers.h
@@ -20,7 +20,7 @@
  * is within a range of supported VF numbers (up to maximum number of VFs that
  * driver can support, including VF0 that represents the PF itself).
  *
- * Note: Effective only on debug builds. See `Xe ASSERTs`_ for more information.
+ * Note: Effective only on debug builds. See `Xe Asserts`_ for more information.
  */
 #define xe_sriov_pf_assert_vfid(xe, vfid) \
 	xe_assert((xe), (vfid) <= xe_sriov_pf_get_totalvfs(xe))
diff --git a/drivers/gpu/drm/xe/xe_sriov_types.h b/drivers/gpu/drm/xe/xe_sriov_types.h
index c7b7ad4af5c8..ca94382a721e 100644
--- a/drivers/gpu/drm/xe/xe_sriov_types.h
+++ b/drivers/gpu/drm/xe/xe_sriov_types.h
@@ -9,6 +9,7 @@
 #include <linux/build_bug.h>
 #include <linux/mutex.h>
 #include <linux/types.h>
+#include <linux/workqueue_types.h>
 
 /**
  * VFID - Virtual Function Identifier
@@ -56,4 +57,20 @@ struct xe_device_pf {
 	struct mutex master_lock;
 };
 
+/**
+ * struct xe_device_vf - Xe Virtual Function related data
+ *
+ * The data in this structure is valid only if driver is running in the
+ * @XE_SRIOV_MODE_VF mode.
+ */
+struct xe_device_vf {
+	/** @migration: VF Migration state data */
+	struct {
+		/** @migration.worker: VF migration recovery worker */
+		struct work_struct worker;
+		/** @migration.gt_flags: Per-GT request flags for VF migration recovery */
+		unsigned long gt_flags;
+	} migration;
+};
+
 #endif
diff --git a/drivers/gpu/drm/xe/xe_sriov_vf.c b/drivers/gpu/drm/xe/xe_sriov_vf.c
new file mode 100644
index 000000000000..c1275e64aa9c
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_sriov_vf.c
@@ -0,0 +1,263 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2023-2024 Intel Corporation
+ */
+
+#include <drm/drm_managed.h>
+
+#include "xe_assert.h"
+#include "xe_device.h"
+#include "xe_gt_sriov_printk.h"
+#include "xe_gt_sriov_vf.h"
+#include "xe_pm.h"
+#include "xe_sriov.h"
+#include "xe_sriov_printk.h"
+#include "xe_sriov_vf.h"
+
+/**
+ * DOC: VF restore procedure in PF KMD and VF KMD
+ *
+ * Restoring previously saved state of a VF is one of core features of
+ * SR-IOV. All major VM Management applications allow saving and restoring
+ * the VM state, and doing that to a VM which uses SRIOV VF as one of
+ * the accessible devices requires support from KMD on both PF and VF side.
+ * VMM initiates all required operations through VFIO module, which then
+ * translates them into PF KMD calls. This description will focus on these
+ * calls, leaving out the module which initiates these steps (VFIO).
+ *
+ * In order to start the restore procedure, GuC needs to keep the VF in
+ * proper state. The PF driver can ensure GuC set it to VF_READY state
+ * by provisioning the VF, which in turn can be done after Function Level
+ * Reset of said VF (or after it was freshly created - in that case FLR
+ * is not needed). The FLR procedure ends with GuC sending message
+ * `GUC_PF_NOTIFY_VF_FLR_DONE`, and then provisioning data is sent to GuC.
+ * After the provisioning is completed, the VF needs to be paused, and
+ * at that point the actual restore can begin.
+ *
+ * During VF Restore, state of several resources is restored. These may
+ * include local memory content (system memory is restored by VMM itself),
+ * values of MMIO registers, stateless compression metadata and others.
+ * The final resource which also needs restoring is state of the VF
+ * submission maintained within GuC. For that, `GUC_PF_OPCODE_VF_RESTORE`
+ * message is used, with reference to the state blob to be consumed by
+ * GuC.
+ *
+ * Next, when VFIO is asked to set the VM into running state, the PF driver
+ * sends `GUC_PF_TRIGGER_VF_RESUME` to GuC. When sent after restore, this
+ * changes VF state within GuC to `VF_RESFIX_BLOCKED` rather than the
+ * usual `VF_RUNNING`. At this point GuC triggers an interrupt to inform
+ * the VF KMD within the VM that it was migrated.
+ *
+ * As soon as Virtual GPU of the VM starts, the VF driver within receives
+ * the MIGRATED interrupt and schedules post-migration recovery worker.
+ * That worker queries GuC for new provisioning (using MMIO communication),
+ * and applies fixups to any non-virtualized resources used by the VF.
+ *
+ * When the VF driver is ready to continue operation on the newly connected
+ * hardware, it sends `VF2GUC_NOTIFY_RESFIX_DONE` which causes it to
+ * enter the long awaited `VF_RUNNING` state, and therefore start handling
+ * CTB messages and scheduling workloads from the VF::
+ *
+ *      PF                             GuC                              VF
+ *     [ ]                              |                               |
+ *     [ ] PF2GUC_VF_CONTROL(pause)     |                               |
+ *     [ ]---------------------------> [ ]                              |
+ *     [ ]                             [ ]  GuC sets new VF state to    |
+ *     [ ]                             [ ]------- VF_READY_PAUSED       |
+ *     [ ]                             [ ]      |                       |
+ *     [ ]                             [ ] <-----                       |
+ *     [ ] success                     [ ]                              |
+ *     [ ] <---------------------------[ ]                              |
+ *     [ ]                              |                               |
+ *     [ ] PF loads resources from the  |                               |
+ *     [ ]------- saved image supplied  |                               |
+ *     [ ]      |                       |                               |
+ *     [ ] <-----                       |                               |
+ *     [ ]                              |                               |
+ *     [ ] GUC_PF_OPCODE_VF_RESTORE     |                               |
+ *     [ ]---------------------------> [ ]                              |
+ *     [ ]                             [ ]  GuC loads contexts and CTB  |
+ *     [ ]                             [ ]------- state from image      |
+ *     [ ]                             [ ]      |                       |
+ *     [ ]                             [ ] <-----                       |
+ *     [ ]                             [ ]                              |
+ *     [ ]                             [ ]  GuC sets new VF state to    |
+ *     [ ]                             [ ]------- VF_RESFIX_PAUSED      |
+ *     [ ]                             [ ]      |                       |
+ *     [ ] success                     [ ] <-----                       |
+ *     [ ] <---------------------------[ ]                              |
+ *     [ ]                              |                               |
+ *     [ ] GUC_PF_TRIGGER_VF_RESUME     |                               |
+ *     [ ]---------------------------> [ ]                              |
+ *     [ ]                             [ ]  GuC sets new VF state to    |
+ *     [ ]                             [ ]------- VF_RESFIX_BLOCKED     |
+ *     [ ]                             [ ]      |                       |
+ *     [ ]                             [ ] <-----                       |
+ *     [ ]                             [ ]                              |
+ *     [ ]                             [ ] GUC_INTR_SW_INT_0            |
+ *     [ ] success                     [ ]---------------------------> [ ]
+ *     [ ] <---------------------------[ ]                             [ ]
+ *      |                               |      VF2GUC_QUERY_SINGLE_KLV [ ]
+ *      |                              [ ] <---------------------------[ ]
+ *      |                              [ ]                             [ ]
+ *      |                              [ ]        new VF provisioning  [ ]
+ *      |                              [ ]---------------------------> [ ]
+ *      |                               |                              [ ]
+ *      |                               |       VF driver applies post [ ]
+ *      |                               |      migration fixups -------[ ]
+ *      |                               |                       |      [ ]
+ *      |                               |                       -----> [ ]
+ *      |                               |                              [ ]
+ *      |                               |    VF2GUC_NOTIFY_RESFIX_DONE [ ]
+ *      |                              [ ] <---------------------------[ ]
+ *      |                              [ ]                             [ ]
+ *      |                              [ ]  GuC sets new VF state to   [ ]
+ *      |                              [ ]------- VF_RUNNING           [ ]
+ *      |                              [ ]      |                      [ ]
+ *      |                              [ ] <-----                      [ ]
+ *      |                              [ ]                     success [ ]
+ *      |                              [ ]---------------------------> [ ]
+ *      |                               |                               |
+ *      |                               |                               |
+ */
+
+static void migration_worker_func(struct work_struct *w);
+
+/**
+ * xe_sriov_vf_init_early - Initialize SR-IOV VF specific data.
+ * @xe: the &xe_device to initialize
+ */
+void xe_sriov_vf_init_early(struct xe_device *xe)
+{
+	INIT_WORK(&xe->sriov.vf.migration.worker, migration_worker_func);
+}
+
+/**
+ * vf_post_migration_requery_guc - Re-query GuC for current VF provisioning.
+ * @xe: the &xe_device struct instance
+ *
+ * After migration, we need to re-query all VF configuration to make sure
+ * they match previous provisioning. Note that most of VF provisioning
+ * shall be the same, except GGTT range, since GGTT is not virtualized per-VF.
+ *
+ * Returns: 0 if the operation completed successfully, or a negative error
+ * code otherwise.
+ */
+static int vf_post_migration_requery_guc(struct xe_device *xe)
+{
+	struct xe_gt *gt;
+	unsigned int id;
+	int err, ret = 0;
+
+	for_each_gt(gt, xe, id) {
+		err = xe_gt_sriov_vf_query_config(gt);
+		ret = ret ?: err;
+	}
+
+	return ret;
+}
+
+/*
+ * vf_post_migration_imminent - Check if post-restore recovery is coming.
+ * @xe: the &xe_device struct instance
+ *
+ * Return: True if migration recovery worker will soon be running. Any worker currently
+ * executing does not affect the result.
+ */
+static bool vf_post_migration_imminent(struct xe_device *xe)
+{
+	return xe->sriov.vf.migration.gt_flags != 0 ||
+	work_pending(&xe->sriov.vf.migration.worker);
+}
+
+/*
+ * Notify all GuCs about resource fixups apply finished.
+ */
+static void vf_post_migration_notify_resfix_done(struct xe_device *xe)
+{
+	struct xe_gt *gt;
+	unsigned int id;
+
+	for_each_gt(gt, xe, id) {
+		if (vf_post_migration_imminent(xe))
+			goto skip;
+		xe_gt_sriov_vf_notify_resfix_done(gt);
+	}
+	return;
+
+skip:
+	drm_dbg(&xe->drm, "another recovery imminent, skipping notifications\n");
+}
+
+static void vf_post_migration_recovery(struct xe_device *xe)
+{
+	int err;
+
+	drm_dbg(&xe->drm, "migration recovery in progress\n");
+	xe_pm_runtime_get(xe);
+	err = vf_post_migration_requery_guc(xe);
+	if (vf_post_migration_imminent(xe))
+		goto defer;
+	if (unlikely(err))
+		goto fail;
+
+	/* FIXME: add the recovery steps */
+	vf_post_migration_notify_resfix_done(xe);
+	xe_pm_runtime_put(xe);
+	drm_notice(&xe->drm, "migration recovery ended\n");
+	return;
+defer:
+	xe_pm_runtime_put(xe);
+	drm_dbg(&xe->drm, "migration recovery deferred\n");
+	return;
+fail:
+	xe_pm_runtime_put(xe);
+	drm_err(&xe->drm, "migration recovery failed (%pe)\n", ERR_PTR(err));
+	xe_device_declare_wedged(xe);
+}
+
+static void migration_worker_func(struct work_struct *w)
+{
+	struct xe_device *xe = container_of(w, struct xe_device,
+					    sriov.vf.migration.worker);
+
+	vf_post_migration_recovery(xe);
+}
+
+static bool vf_ready_to_recovery_on_all_gts(struct xe_device *xe)
+{
+	struct xe_gt *gt;
+	unsigned int id;
+
+	for_each_gt(gt, xe, id) {
+		if (!test_bit(id, &xe->sriov.vf.migration.gt_flags)) {
+			xe_gt_sriov_dbg_verbose(gt, "still not ready to recover\n");
+			return false;
+		}
+	}
+	return true;
+}
+
+/**
+ * xe_sriov_vf_start_migration_recovery - Start VF migration recovery.
+ * @xe: the &xe_device to start recovery on
+ *
+ * This function shall be called only by VF.
+ */
+void xe_sriov_vf_start_migration_recovery(struct xe_device *xe)
+{
+	bool started;
+
+	xe_assert(xe, IS_SRIOV_VF(xe));
+
+	if (!vf_ready_to_recovery_on_all_gts(xe))
+		return;
+
+	WRITE_ONCE(xe->sriov.vf.migration.gt_flags, 0);
+	/* Ensure other threads see that no flags are set now. */
+	smp_mb();
+
+	started = queue_work(xe->sriov.wq, &xe->sriov.vf.migration.worker);
+	drm_info(&xe->drm, "VF migration recovery %s\n", started ?
+		 "scheduled" : "already in progress");
+}
diff --git a/drivers/gpu/drm/xe/xe_sriov_vf.h b/drivers/gpu/drm/xe/xe_sriov_vf.h
new file mode 100644
index 000000000000..7b8622cff2b7
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_sriov_vf.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2023-2024 Intel Corporation
+ */
+
+#ifndef _XE_SRIOV_VF_H_
+#define _XE_SRIOV_VF_H_
+
+struct xe_device;
+
+void xe_sriov_vf_init_early(struct xe_device *xe);
+void xe_sriov_vf_start_migration_recovery(struct xe_device *xe);
+
+#endif
diff --git a/drivers/gpu/drm/xe/xe_trace.h b/drivers/gpu/drm/xe/xe_trace.h
index 91130ad8999c..d5281de04d54 100644
--- a/drivers/gpu/drm/xe/xe_trace.h
+++ b/drivers/gpu/drm/xe/xe_trace.h
@@ -211,6 +211,7 @@ DECLARE_EVENT_CLASS(xe_sched_job,
 			     __string(dev, __dev_name_eq(job->q))
 			     __field(u32, seqno)
 			     __field(u32, lrc_seqno)
+			     __field(u8, gt_id)
 			     __field(u16, guc_id)
 			     __field(u32, guc_state)
 			     __field(u32, flags)
@@ -223,6 +224,7 @@ DECLARE_EVENT_CLASS(xe_sched_job,
 			   __assign_str(dev);
 			   __entry->seqno = xe_sched_job_seqno(job);
 			   __entry->lrc_seqno = xe_sched_job_lrc_seqno(job);
+			   __entry->gt_id = job->q->gt->info.id;
 			   __entry->guc_id = job->q->guc->id;
 			   __entry->guc_state =
 			   atomic_read(&job->q->guc->state);
@@ -232,9 +234,9 @@ DECLARE_EVENT_CLASS(xe_sched_job,
 			   __entry->batch_addr = (u64)job->ptrs[0].batch_addr;
 			   ),
 
-		    TP_printk("dev=%s, fence=%p, seqno=%u, lrc_seqno=%u, guc_id=%d, batch_addr=0x%012llx, guc_state=0x%x, flags=0x%x, error=%d",
+		    TP_printk("dev=%s, fence=%p, seqno=%u, lrc_seqno=%u, gt=%u, guc_id=%d, batch_addr=0x%012llx, guc_state=0x%x, flags=0x%x, error=%d",
 			      __get_str(dev), __entry->fence, __entry->seqno,
-			      __entry->lrc_seqno, __entry->guc_id,
+			      __entry->lrc_seqno, __entry->gt_id, __entry->guc_id,
 			      __entry->batch_addr, __entry->guc_state,
 			      __entry->flags, __entry->error)
 );
@@ -282,6 +284,7 @@ DECLARE_EVENT_CLASS(xe_sched_msg,
 			     __string(dev, __dev_name_eq(((struct xe_exec_queue *)msg->private_data)))
 			     __field(u32, opcode)
 			     __field(u16, guc_id)
+			     __field(u8, gt_id)
 			     ),
 
 		    TP_fast_assign(
@@ -289,9 +292,11 @@ DECLARE_EVENT_CLASS(xe_sched_msg,
 			   __entry->opcode = msg->opcode;
 			   __entry->guc_id =
 			   ((struct xe_exec_queue *)msg->private_data)->guc->id;
+			   __entry->gt_id =
+			   ((struct xe_exec_queue *)msg->private_data)->gt->info.id;
 			   ),
 
-		    TP_printk("dev=%s, guc_id=%d, opcode=%u", __get_str(dev), __entry->guc_id,
+		    TP_printk("dev=%s, gt=%u guc_id=%d, opcode=%u", __get_str(dev), __entry->gt_id, __entry->guc_id,
 			      __entry->opcode)
 );
 
diff --git a/drivers/gpu/drm/xe/xe_trace_bo.h b/drivers/gpu/drm/xe/xe_trace_bo.h
index 30a3cfbaaa09..1762dd30ba6d 100644
--- a/drivers/gpu/drm/xe/xe_trace_bo.h
+++ b/drivers/gpu/drm/xe/xe_trace_bo.h
@@ -48,6 +48,11 @@ DEFINE_EVENT(xe_bo, xe_bo_cpu_fault,
 	     TP_ARGS(bo)
 );
 
+DEFINE_EVENT(xe_bo, xe_bo_validate,
+	     TP_PROTO(struct xe_bo *bo),
+	     TP_ARGS(bo)
+);
+
 TRACE_EVENT(xe_bo_move,
 	    TP_PROTO(struct xe_bo *bo, uint32_t new_placement, uint32_t old_placement,
 		     bool move_lacks_source),
diff --git a/drivers/gpu/drm/xe/xe_trace_lrc.c b/drivers/gpu/drm/xe/xe_trace_lrc.c
new file mode 100644
index 000000000000..ab9b7e2970bc
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_trace_lrc.c
@@ -0,0 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright © 2024 Intel Corporation
+ */
+
+#ifndef __CHECKER__
+#define CREATE_TRACE_POINTS
+#include "xe_trace_lrc.h"
+#endif
diff --git a/drivers/gpu/drm/xe/xe_trace_lrc.h b/drivers/gpu/drm/xe/xe_trace_lrc.h
new file mode 100644
index 000000000000..5c669a0b2180
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_trace_lrc.h
@@ -0,0 +1,52 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright © 2024 Intel Corporation
+ */
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM xe
+
+#if !defined(_XE_TRACE_LRC_H_) || defined(TRACE_HEADER_MULTI_READ)
+#define _XE_TRACE_LRC_H_
+
+#include <linux/tracepoint.h>
+#include <linux/types.h>
+
+#include "xe_gt_types.h"
+#include "xe_lrc.h"
+#include "xe_lrc_types.h"
+
+#define __dev_name_lrc(lrc)	dev_name(gt_to_xe((lrc)->fence_ctx.gt)->drm.dev)
+
+TRACE_EVENT(xe_lrc_update_timestamp,
+	    TP_PROTO(struct xe_lrc *lrc, uint32_t old),
+	    TP_ARGS(lrc, old),
+	    TP_STRUCT__entry(
+		     __field(struct xe_lrc *, lrc)
+		     __field(u32, old)
+		     __field(u32, new)
+		     __string(name, lrc->fence_ctx.name)
+		     __string(device_id, __dev_name_lrc(lrc))
+	    ),
+
+	    TP_fast_assign(
+		   __entry->lrc	= lrc;
+		   __entry->old = old;
+		   __entry->new = lrc->ctx_timestamp;
+		   __assign_str(name);
+		   __assign_str(device_id);
+		   ),
+	    TP_printk("lrc=:%p lrc->name=%s old=%u new=%u device_id:%s",
+		      __entry->lrc, __get_str(name),
+		      __entry->old, __entry->new,
+		      __get_str(device_id))
+);
+
+#endif
+
+/* This part must be outside protection */
+#undef TRACE_INCLUDE_PATH
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_PATH ../../drivers/gpu/drm/xe
+#define TRACE_INCLUDE_FILE xe_trace_lrc
+#include <trace/define_trace.h>
diff --git a/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c b/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
index 423b261ea743..c95728c45ea4 100644
--- a/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
+++ b/drivers/gpu/drm/xe/xe_ttm_vram_mgr.c
@@ -52,7 +52,7 @@ static int xe_ttm_vram_mgr_new(struct ttm_resource_manager *man,
 	struct xe_ttm_vram_mgr *mgr = to_xe_ttm_vram_mgr(man);
 	struct xe_ttm_vram_mgr_resource *vres;
 	struct drm_buddy *mm = &mgr->mm;
-	u64 size, remaining_size, min_page_size;
+	u64 size, min_page_size;
 	unsigned long lpfn;
 	int err;
 
@@ -98,17 +98,6 @@ static int xe_ttm_vram_mgr_new(struct ttm_resource_manager *man,
 		goto error_fini;
 	}
 
-	if (WARN_ON(min_page_size > SZ_2G)) { /* FIXME: sg limit */
-		err = -EINVAL;
-		goto error_fini;
-	}
-
-	if (WARN_ON((size > SZ_2G &&
-		     (vres->base.placement & TTM_PL_FLAG_CONTIGUOUS)))) {
-		err = -EINVAL;
-		goto error_fini;
-	}
-
 	if (WARN_ON(!IS_ALIGNED(size, min_page_size))) {
 		err = -EINVAL;
 		goto error_fini;
@@ -116,12 +105,11 @@ static int xe_ttm_vram_mgr_new(struct ttm_resource_manager *man,
 
 	mutex_lock(&mgr->lock);
 	if (lpfn <= mgr->visible_size >> PAGE_SHIFT && size > mgr->visible_avail) {
-		mutex_unlock(&mgr->lock);
 		err = -ENOSPC;
-		goto error_fini;
+		goto error_unlock;
 	}
 
-	if (place->fpfn + (size >> PAGE_SHIFT) != place->lpfn &&
+	if (place->fpfn + (size >> PAGE_SHIFT) != lpfn &&
 	    place->flags & TTM_PL_FLAG_CONTIGUOUS) {
 		size = roundup_pow_of_two(size);
 		min_page_size = size;
@@ -129,25 +117,11 @@ static int xe_ttm_vram_mgr_new(struct ttm_resource_manager *man,
 		lpfn = max_t(unsigned long, place->fpfn + (size >> PAGE_SHIFT), lpfn);
 	}
 
-	remaining_size = size;
-	do {
-		/*
-		 * Limit maximum size to 2GiB due to SG table limitations.
-		 * FIXME: Should maybe be handled as part of sg construction.
-		 */
-		u64 alloc_size = min_t(u64, remaining_size, SZ_2G);
-
-		err = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
-					     (u64)lpfn << PAGE_SHIFT,
-					     alloc_size,
-					     min_page_size,
-					     &vres->blocks,
-					     vres->flags);
-		if (err)
-			goto error_free_blocks;
-
-		remaining_size -= alloc_size;
-	} while (remaining_size);
+	err = drm_buddy_alloc_blocks(mm, (u64)place->fpfn << PAGE_SHIFT,
+				     (u64)lpfn << PAGE_SHIFT, size,
+				     min_page_size, &vres->blocks, vres->flags);
+	if (err)
+		goto error_unlock;
 
 	if (place->flags & TTM_PL_FLAG_CONTIGUOUS) {
 		if (!drm_buddy_block_trim(mm, NULL, vres->base.size, &vres->blocks))
@@ -194,9 +168,7 @@ static int xe_ttm_vram_mgr_new(struct ttm_resource_manager *man,
 
 	*res = &vres->base;
 	return 0;
-
-error_free_blocks:
-	drm_buddy_free_list(mm, &vres->blocks, 0);
+error_unlock:
 	mutex_unlock(&mgr->lock);
 error_fini:
 	ttm_resource_fini(man, &vres->base);
@@ -393,7 +365,8 @@ int xe_ttm_vram_mgr_alloc_sgt(struct xe_device *xe,
 	xe_res_first(res, offset, length, &cursor);
 	while (cursor.remaining) {
 		num_entries++;
-		xe_res_next(&cursor, cursor.size);
+		/* Limit maximum size to 2GiB due to SG table limitations. */
+		xe_res_next(&cursor, min_t(u64, cursor.size, SZ_2G));
 	}
 
 	r = sg_alloc_table(*sgt, num_entries, GFP_KERNEL);
@@ -413,7 +386,7 @@ int xe_ttm_vram_mgr_alloc_sgt(struct xe_device *xe,
 	xe_res_first(res, offset, length, &cursor);
 	for_each_sgtable_sg((*sgt), sg, i) {
 		phys_addr_t phys = cursor.start + tile->mem.vram.io_start;
-		size_t size = cursor.size;
+		size_t size = min_t(u64, cursor.size, SZ_2G);
 		dma_addr_t addr;
 
 		addr = dma_map_resource(dev, phys, size, dir,
@@ -426,7 +399,7 @@ int xe_ttm_vram_mgr_alloc_sgt(struct xe_device *xe,
 		sg_dma_address(sg) = addr;
 		sg_dma_len(sg) = size;
 
-		xe_res_next(&cursor, cursor.size);
+		xe_res_next(&cursor, size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
index 00ea57c2f4b9..b4a44e1ea167 100644
--- a/drivers/gpu/drm/xe/xe_vm.c
+++ b/drivers/gpu/drm/xe/xe_vm.c
@@ -732,13 +732,14 @@ static int xe_vma_ops_alloc(struct xe_vma_ops *vops, bool array_of_binds)
 		vops->pt_update_ops[i].ops =
 			kmalloc_array(vops->pt_update_ops[i].num_ops,
 				      sizeof(*vops->pt_update_ops[i].ops),
-				      GFP_KERNEL);
+				      GFP_KERNEL | __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
 		if (!vops->pt_update_ops[i].ops)
 			return array_of_binds ? -ENOBUFS : -ENOMEM;
 	}
 
 	return 0;
 }
+ALLOW_ERROR_INJECTION(xe_vma_ops_alloc, ERRNO);
 
 static void xe_vma_ops_fini(struct xe_vma_ops *vops)
 {
@@ -1351,6 +1352,7 @@ static int xe_vm_create_scratch(struct xe_device *xe, struct xe_tile *tile,
 
 	return 0;
 }
+ALLOW_ERROR_INJECTION(xe_vm_create_scratch, ERRNO);
 
 static void xe_vm_free_scratch(struct xe_vm *vm)
 {
@@ -1977,6 +1979,7 @@ vm_bind_ioctl_ops_create(struct xe_vm *vm, struct xe_bo *bo,
 
 	return ops;
 }
+ALLOW_ERROR_INJECTION(vm_bind_ioctl_ops_create, ERRNO);
 
 static struct xe_vma *new_vma(struct xe_vm *vm, struct drm_gpuva_op_map *op,
 			      u16 pat_index, unsigned int flags)
@@ -2356,13 +2359,15 @@ static int vma_lock_and_validate(struct drm_exec *exec, struct xe_vma *vma,
 				 bool validate)
 {
 	struct xe_bo *bo = xe_vma_bo(vma);
+	struct xe_vm *vm = xe_vma_vm(vma);
 	int err = 0;
 
 	if (bo) {
 		if (!bo->vm)
 			err = drm_exec_lock_obj(exec, &bo->ttm.base);
 		if (!err && validate)
-			err = xe_bo_validate(bo, xe_vma_vm(vma), true);
+			err = xe_bo_validate(bo, vm,
+					     !xe_vm_in_preempt_fence_mode(vm));
 	}
 
 	return err;
@@ -2696,6 +2701,7 @@ unlock:
 	drm_exec_fini(&exec);
 	return err;
 }
+ALLOW_ERROR_INJECTION(vm_bind_ioctl_ops_execute, ERRNO);
 
 #define SUPPORTED_FLAGS_STUB  \
 	(DRM_XE_VM_BIND_FLAG_READONLY | \
@@ -2732,7 +2738,8 @@ static int vm_bind_ioctl_check_args(struct xe_device *xe,
 
 		*bind_ops = kvmalloc_array(args->num_binds,
 					   sizeof(struct drm_xe_vm_bind_op),
-					   GFP_KERNEL | __GFP_ACCOUNT);
+					   GFP_KERNEL | __GFP_ACCOUNT |
+					   __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
 		if (!*bind_ops)
 			return args->num_binds > 1 ? -ENOBUFS : -ENOMEM;
 
@@ -2972,14 +2979,16 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 
 	if (args->num_binds) {
 		bos = kvcalloc(args->num_binds, sizeof(*bos),
-			       GFP_KERNEL | __GFP_ACCOUNT);
+			       GFP_KERNEL | __GFP_ACCOUNT |
+			       __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
 		if (!bos) {
 			err = -ENOMEM;
 			goto release_vm_lock;
 		}
 
 		ops = kvcalloc(args->num_binds, sizeof(*ops),
-			       GFP_KERNEL | __GFP_ACCOUNT);
+			       GFP_KERNEL | __GFP_ACCOUNT |
+			       __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
 		if (!ops) {
 			err = -ENOMEM;
 			goto release_vm_lock;
@@ -3302,7 +3311,6 @@ void xe_vm_snapshot_capture_delayed(struct xe_vm_snapshot *snap)
 
 	for (int i = 0; i < snap->num_snaps; i++) {
 		struct xe_bo *bo = snap->snap[i].bo;
-		struct iosys_map src;
 		int err;
 
 		if (IS_ERR(snap->snap[i].data))
@@ -3315,16 +3323,8 @@ void xe_vm_snapshot_capture_delayed(struct xe_vm_snapshot *snap)
 		}
 
 		if (bo) {
-			xe_bo_lock(bo, false);
-			err = ttm_bo_vmap(&bo->ttm, &src);
-			if (!err) {
-				xe_map_memcpy_from(xe_bo_device(bo),
-						   snap->snap[i].data,
-						   &src, snap->snap[i].bo_ofs,
-						   snap->snap[i].len);
-				ttm_bo_vunmap(&bo->ttm, &src);
-			}
-			xe_bo_unlock(bo);
+			err = xe_bo_read(bo, snap->snap[i].bo_ofs,
+					 snap->snap[i].data, snap->snap[i].len);
 		} else {
 			void __user *userptr = (void __user *)(size_t)snap->snap[i].bo_ofs;
 
diff --git a/drivers/gpu/drm/xe/xe_vm_doc.h b/drivers/gpu/drm/xe/xe_vm_doc.h
index 4d33f310b653..078786958403 100644
--- a/drivers/gpu/drm/xe/xe_vm_doc.h
+++ b/drivers/gpu/drm/xe/xe_vm_doc.h
@@ -64,8 +64,8 @@
  *	update page level 2 PDE[1] to page level 3b phys address (GPU)
  *
  *	bind BO2 0x1ff000-0x201000
- *	update page level 3a PTE[511] to BO2 phys addres (GPU)
- *	update page level 3b PTE[0] to BO2 phys addres + 0x1000 (GPU)
+ *	update page level 3a PTE[511] to BO2 phys address (GPU)
+ *	update page level 3b PTE[0] to BO2 phys address + 0x1000 (GPU)
  *
  * GPU bypass
  * ~~~~~~~~~~
@@ -192,7 +192,7 @@
  *
  * If a VM is in fault mode (TODO: link to fault mode), new bind operations that
  * create mappings are by default deferred to the page fault handler (first
- * use). This behavior can be overriden by setting the flag
+ * use). This behavior can be overridden by setting the flag
  * DRM_XE_VM_BIND_FLAG_IMMEDIATE which indicates to creating the mapping
  * immediately.
  *
@@ -209,7 +209,7 @@
  *
  * Since this a core kernel managed memory the kernel can move this memory
  * whenever it wants. We register an invalidation MMU notifier to alert XE when
- * a user poiter is about to move. The invalidation notifier needs to block
+ * a user pointer is about to move. The invalidation notifier needs to block
  * until all pending users (jobs or compute mode engines) of the userptr are
  * idle to ensure no faults. This done by waiting on all of VM's dma-resv slots.
  *
@@ -252,7 +252,7 @@
  * Rebind worker
  * -------------
  *
- * The rebind worker is very similar to an exec. It is resposible for rebinding
+ * The rebind worker is very similar to an exec. It is responsible for rebinding
  * evicted BOs or userptrs, waiting on those operations, installing new preempt
  * fences, and finally resuming executing of engines in the VM.
  *
@@ -317,11 +317,11 @@
  * are not allowed, only long running workloads and ULLS are enabled on a faulting
  * VM.
  *
- * Defered VM binds
+ * Deferred VM binds
  * ----------------
  *
  * By default, on a faulting VM binds just allocate the VMA and the actual
- * updating of the page tables is defered to the page fault handler. This
+ * updating of the page tables is deferred to the page fault handler. This
  * behavior can be overridden by setting the flag DRM_XE_VM_BIND_FLAG_IMMEDIATE in
  * the VM bind which will then do the bind immediately.
  *
@@ -500,18 +500,18 @@
  * Slot waiting
  * ------------
  *
- * 1. The exection of all jobs from kernel ops shall wait on all slots
+ * 1. The execution of all jobs from kernel ops shall wait on all slots
  * (DMA_RESV_USAGE_PREEMPT_FENCE) of either an external BO or VM (depends on if
  * kernel op is operating on external or private BO)
  *
- * 2. In non-compute mode, the exection of all jobs from rebinds in execs shall
+ * 2. In non-compute mode, the execution of all jobs from rebinds in execs shall
  * wait on the DMA_RESV_USAGE_KERNEL slot of either an external BO or VM
  * (depends on if the rebind is operatiing on an external or private BO)
  *
- * 3. In non-compute mode, the exection of all jobs from execs shall wait on the
+ * 3. In non-compute mode, the execution of all jobs from execs shall wait on the
  * last rebind job
  *
- * 4. In compute mode, the exection of all jobs from rebinds in the rebind
+ * 4. In compute mode, the execution of all jobs from rebinds in the rebind
  * worker shall wait on the DMA_RESV_USAGE_KERNEL slot of either an external BO
  * or VM (depends on if rebind is operating on external or private BO)
  *
diff --git a/drivers/gpu/drm/xe/xe_vsec.c b/drivers/gpu/drm/xe/xe_vsec.c
new file mode 100644
index 000000000000..b378848d3b7b
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_vsec.c
@@ -0,0 +1,233 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright © 2024 Intel Corporation */
+#include <linux/bitfield.h>
+#include <linux/bits.h>
+#include <linux/cleanup.h>
+#include <linux/errno.h>
+#include <linux/intel_vsec.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/pci.h>
+#include <linux/types.h>
+
+#include "xe_device.h"
+#include "xe_device_types.h"
+#include "xe_drv.h"
+#include "xe_mmio.h"
+#include "xe_platform_types.h"
+#include "xe_pm.h"
+#include "xe_vsec.h"
+
+#include "regs/xe_pmt.h"
+
+/* PMT GUID value for BMG devices.  NOTE: this is NOT a PCI id */
+#define BMG_DEVICE_ID 0xE2F8
+
+static struct intel_vsec_header bmg_telemetry = {
+	.length = 0x10,
+	.id = VSEC_ID_TELEMETRY,
+	.num_entries = 2,
+	.entry_size = 4,
+	.tbir = 0,
+	.offset = BMG_DISCOVERY_OFFSET,
+};
+
+static struct intel_vsec_header bmg_punit_crashlog = {
+	.length = 0x10,
+	.id = VSEC_ID_CRASHLOG,
+	.num_entries = 1,
+	.entry_size = 4,
+	.tbir = 0,
+	.offset = BMG_DISCOVERY_OFFSET + 0x60,
+};
+
+static struct intel_vsec_header bmg_oobmsm_crashlog = {
+	.length = 0x10,
+	.id = VSEC_ID_CRASHLOG,
+	.num_entries = 1,
+	.entry_size = 4,
+	.tbir = 0,
+	.offset = BMG_DISCOVERY_OFFSET + 0x78,
+};
+
+static struct intel_vsec_header *bmg_capabilities[] = {
+	&bmg_telemetry,
+	&bmg_punit_crashlog,
+	&bmg_oobmsm_crashlog,
+	NULL
+};
+
+enum xe_vsec {
+	XE_VSEC_UNKNOWN = 0,
+	XE_VSEC_BMG,
+};
+
+static struct intel_vsec_platform_info xe_vsec_info[] = {
+	[XE_VSEC_BMG] = {
+		.caps = VSEC_CAP_TELEMETRY | VSEC_CAP_CRASHLOG,
+		.headers = bmg_capabilities,
+	},
+	{ }
+};
+
+/*
+ * The GUID will have the following bits to decode:
+ *   [0:3]   - {Telemetry space iteration number (0,1,..)}
+ *   [4:7]   - Segment (SEGMENT_INDEPENDENT-0, Client-1, Server-2)
+ *   [8:11]  - SOC_SKU
+ *   [12:27] – Device ID – changes for each down bin SKU’s
+ *   [28:29] - Capability Type (Crashlog-0, Telemetry Aggregator-1, Watcher-2)
+ *   [30:31] - Record-ID (0-PUNIT, 1-OOBMSM_0, 2-OOBMSM_1)
+ */
+#define GUID_TELEM_ITERATION	GENMASK(3, 0)
+#define GUID_SEGMENT		GENMASK(7, 4)
+#define GUID_SOC_SKU		GENMASK(11, 8)
+#define GUID_DEVICE_ID		GENMASK(27, 12)
+#define GUID_CAP_TYPE		GENMASK(29, 28)
+#define GUID_RECORD_ID		GENMASK(31, 30)
+
+#define PUNIT_TELEMETRY_OFFSET		0x0200
+#define PUNIT_WATCHER_OFFSET		0x14A0
+#define OOBMSM_0_WATCHER_OFFSET		0x18D8
+#define OOBMSM_1_TELEMETRY_OFFSET	0x1000
+
+enum record_id {
+	PUNIT,
+	OOBMSM_0,
+	OOBMSM_1,
+};
+
+enum capability {
+	CRASHLOG,
+	TELEMETRY,
+	WATCHER,
+};
+
+static int xe_guid_decode(u32 guid, int *index, u32 *offset)
+{
+	u32 record_id = FIELD_GET(GUID_RECORD_ID, guid);
+	u32 cap_type  = FIELD_GET(GUID_CAP_TYPE, guid);
+	u32 device_id = FIELD_GET(GUID_DEVICE_ID, guid);
+
+	if (device_id != BMG_DEVICE_ID)
+		return -ENODEV;
+
+	if (cap_type > WATCHER)
+		return -EINVAL;
+
+	*offset = 0;
+
+	if (cap_type == CRASHLOG) {
+		*index = record_id == PUNIT ? 2 : 4;
+		return 0;
+	}
+
+	switch (record_id) {
+	case PUNIT:
+		*index = 0;
+		if (cap_type == TELEMETRY)
+			*offset = PUNIT_TELEMETRY_OFFSET;
+		else
+			*offset = PUNIT_WATCHER_OFFSET;
+		break;
+
+	case OOBMSM_0:
+		*index = 1;
+		if (cap_type == WATCHER)
+			*offset = OOBMSM_0_WATCHER_OFFSET;
+		break;
+
+	case OOBMSM_1:
+		*index = 1;
+		if (cap_type == TELEMETRY)
+			*offset = OOBMSM_1_TELEMETRY_OFFSET;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int xe_pmt_telem_read(struct pci_dev *pdev, u32 guid, u64 *data, loff_t user_offset,
+			     u32 count)
+{
+	struct xe_device *xe = pdev_to_xe_device(pdev);
+	void __iomem *telem_addr = xe->mmio.regs + BMG_TELEMETRY_OFFSET;
+	u32 mem_region;
+	u32 offset;
+	int ret;
+
+	ret = xe_guid_decode(guid, &mem_region, &offset);
+	if (ret)
+		return ret;
+
+	telem_addr += offset + user_offset;
+
+	guard(mutex)(&xe->pmt.lock);
+
+	/* indicate that we are not at an appropriate power level */
+	if (!xe_pm_runtime_get_if_active(xe))
+		return -ENODATA;
+
+	/* set SoC re-mapper index register based on GUID memory region */
+	xe_mmio_rmw32(xe_root_tile_mmio(xe), SG_REMAP_INDEX1, SG_REMAP_BITS,
+		      REG_FIELD_PREP(SG_REMAP_BITS, mem_region));
+
+	memcpy_fromio(data, telem_addr, count);
+	xe_pm_runtime_put(xe);
+
+	return count;
+}
+
+static struct pmt_callbacks xe_pmt_cb = {
+	.read_telem = xe_pmt_telem_read,
+};
+
+static const int vsec_platforms[] = {
+	[XE_BATTLEMAGE] = XE_VSEC_BMG,
+};
+
+static enum xe_vsec get_platform_info(struct xe_device *xe)
+{
+	if (xe->info.platform > XE_BATTLEMAGE)
+		return XE_VSEC_UNKNOWN;
+
+	return vsec_platforms[xe->info.platform];
+}
+
+/**
+ * xe_vsec_init - Initialize resources and add intel_vsec auxiliary
+ * interface
+ * @xe: valid xe instance
+ */
+void xe_vsec_init(struct xe_device *xe)
+{
+	struct intel_vsec_platform_info *info;
+	struct device *dev = xe->drm.dev;
+	struct pci_dev *pdev = to_pci_dev(dev);
+	enum xe_vsec platform;
+
+	platform = get_platform_info(xe);
+	if (platform == XE_VSEC_UNKNOWN)
+		return;
+
+	info = &xe_vsec_info[platform];
+	if (!info->headers)
+		return;
+
+	switch (platform) {
+	case XE_VSEC_BMG:
+		info->priv_data = &xe_pmt_cb;
+		break;
+	default:
+		break;
+	}
+
+	/*
+	 * Register a VSEC. Cleanup is handled using device managed
+	 * resources.
+	 */
+	intel_vsec_register(pdev, info);
+}
+MODULE_IMPORT_NS("INTEL_VSEC");
diff --git a/drivers/gpu/drm/xe/xe_vsec.h b/drivers/gpu/drm/xe/xe_vsec.h
new file mode 100644
index 000000000000..5777c53faec2
--- /dev/null
+++ b/drivers/gpu/drm/xe/xe_vsec.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright © 2024 Intel Corporation */
+
+#ifndef _XE_VSEC_H_
+#define _XE_VSEC_H_
+
+struct xe_device;
+
+void xe_vsec_init(struct xe_device *xe);
+
+#endif
diff --git a/drivers/gpu/drm/xe/xe_wa.c b/drivers/gpu/drm/xe/xe_wa.c
index 02cf647f86d8..570fe0376402 100644
--- a/drivers/gpu/drm/xe/xe_wa.c
+++ b/drivers/gpu/drm/xe/xe_wa.c
@@ -607,6 +607,12 @@ static const struct xe_rtp_entry_sr engine_was[] = {
 		       FUNC(xe_rtp_match_first_render_or_compute)),
 	  XE_RTP_ACTIONS(SET(ROW_CHICKEN4, DISABLE_TDL_PUSH))
 	},
+	{ XE_RTP_NAME("16024792527"),
+	  XE_RTP_RULES(GRAPHICS_VERSION(3000), GRAPHICS_STEP(A0, B0),
+		       FUNC(xe_rtp_match_first_render_or_compute)),
+	  XE_RTP_ACTIONS(FIELD_SET(SAMPLER_MODE, SMP_WAIT_FETCH_MERGING_COUNTER,
+				   SMP_FORCE_128B_OVERFETCH))
+	},
 
 	{}
 };
diff --git a/drivers/gpu/drm/xe/xe_wa_oob.rules b/drivers/gpu/drm/xe/xe_wa_oob.rules
index bcd04464b85e..3ed12a85cc60 100644
--- a/drivers/gpu/drm/xe/xe_wa_oob.rules
+++ b/drivers/gpu/drm/xe/xe_wa_oob.rules
@@ -1,3 +1,4 @@
+1607983814	GRAPHICS_VERSION_RANGE(1200, 1210)
 22012773006	GRAPHICS_VERSION_RANGE(1200, 1250)
 14014475959	GRAPHICS_VERSION_RANGE(1270, 1271), GRAPHICS_STEP(A0, B0)
 		PLATFORM(DG2)
diff --git a/include/drm/intel/xe_pciids.h b/include/drm/intel/xe_pciids.h
new file mode 100644
index 000000000000..16d4b8bb590a
--- /dev/null
+++ b/include/drm/intel/xe_pciids.h
@@ -0,0 +1,235 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2022 Intel Corporation
+ */
+
+#ifndef _XE_PCIIDS_H_
+#define _XE_PCIIDS_H_
+
+/*
+ * Lists below can be turned into initializers for a struct pci_device_id
+ * by defining INTEL_VGA_DEVICE:
+ *
+ * #define INTEL_VGA_DEVICE(id, info) { \
+ *	0x8086, id,			\
+ *	~0, ~0,				\
+ *	0x030000, 0xff0000,		\
+ *	(unsigned long) info }
+ *
+ * And then calling like:
+ *
+ * XE_TGL_12_GT1_IDS(INTEL_VGA_DEVICE, ## __VA_ARGS__)
+ *
+ * To turn them into something else, just provide a different macro passed as
+ * first argument.
+ */
+
+/* TGL */
+#define XE_TGL_GT1_IDS(MACRO__, ...)		\
+	MACRO__(0x9A60, ## __VA_ARGS__),	\
+	MACRO__(0x9A68, ## __VA_ARGS__),	\
+	MACRO__(0x9A70, ## __VA_ARGS__)
+
+#define XE_TGL_GT2_IDS(MACRO__, ...)		\
+	MACRO__(0x9A40, ## __VA_ARGS__),	\
+	MACRO__(0x9A49, ## __VA_ARGS__),	\
+	MACRO__(0x9A59, ## __VA_ARGS__),	\
+	MACRO__(0x9A78, ## __VA_ARGS__),	\
+	MACRO__(0x9AC0, ## __VA_ARGS__),	\
+	MACRO__(0x9AC9, ## __VA_ARGS__),	\
+	MACRO__(0x9AD9, ## __VA_ARGS__),	\
+	MACRO__(0x9AF8, ## __VA_ARGS__)
+
+#define XE_TGL_IDS(MACRO__, ...)		\
+	XE_TGL_GT1_IDS(MACRO__, ## __VA_ARGS__),\
+	XE_TGL_GT2_IDS(MACRO__, ## __VA_ARGS__)
+
+/* RKL */
+#define XE_RKL_IDS(MACRO__, ...)		\
+	MACRO__(0x4C80, ## __VA_ARGS__),	\
+	MACRO__(0x4C8A, ## __VA_ARGS__),	\
+	MACRO__(0x4C8B, ## __VA_ARGS__),	\
+	MACRO__(0x4C8C, ## __VA_ARGS__),	\
+	MACRO__(0x4C90, ## __VA_ARGS__),	\
+	MACRO__(0x4C9A, ## __VA_ARGS__)
+
+/* DG1 */
+#define XE_DG1_IDS(MACRO__, ...)		\
+	MACRO__(0x4905, ## __VA_ARGS__),	\
+	MACRO__(0x4906, ## __VA_ARGS__),	\
+	MACRO__(0x4907, ## __VA_ARGS__),	\
+	MACRO__(0x4908, ## __VA_ARGS__),	\
+	MACRO__(0x4909, ## __VA_ARGS__)
+
+/* ADL-S */
+#define XE_ADLS_IDS(MACRO__, ...)		\
+	MACRO__(0x4680, ## __VA_ARGS__),	\
+	MACRO__(0x4682, ## __VA_ARGS__),	\
+	MACRO__(0x4688, ## __VA_ARGS__),	\
+	MACRO__(0x468A, ## __VA_ARGS__),	\
+	MACRO__(0x468B, ## __VA_ARGS__),	\
+	MACRO__(0x4690, ## __VA_ARGS__),	\
+	MACRO__(0x4692, ## __VA_ARGS__),	\
+	MACRO__(0x4693, ## __VA_ARGS__)
+
+/* ADL-P */
+#define XE_ADLP_IDS(MACRO__, ...)		\
+	MACRO__(0x46A0, ## __VA_ARGS__),	\
+	MACRO__(0x46A1, ## __VA_ARGS__),	\
+	MACRO__(0x46A2, ## __VA_ARGS__),	\
+	MACRO__(0x46A3, ## __VA_ARGS__),	\
+	MACRO__(0x46A6, ## __VA_ARGS__),	\
+	MACRO__(0x46A8, ## __VA_ARGS__),	\
+	MACRO__(0x46AA, ## __VA_ARGS__),	\
+	MACRO__(0x462A, ## __VA_ARGS__),	\
+	MACRO__(0x4626, ## __VA_ARGS__),	\
+	MACRO__(0x4628, ## __VA_ARGS__),	\
+	MACRO__(0x46B0, ## __VA_ARGS__),	\
+	MACRO__(0x46B1, ## __VA_ARGS__),	\
+	MACRO__(0x46B2, ## __VA_ARGS__),	\
+	MACRO__(0x46B3, ## __VA_ARGS__),	\
+	MACRO__(0x46C0, ## __VA_ARGS__),	\
+	MACRO__(0x46C1, ## __VA_ARGS__),	\
+	MACRO__(0x46C2, ## __VA_ARGS__),	\
+	MACRO__(0x46C3, ## __VA_ARGS__)
+
+/* ADL-N */
+#define XE_ADLN_IDS(MACRO__, ...)		\
+	MACRO__(0x46D0, ## __VA_ARGS__),	\
+	MACRO__(0x46D1, ## __VA_ARGS__),	\
+	MACRO__(0x46D2, ## __VA_ARGS__),	\
+	MACRO__(0x46D3, ## __VA_ARGS__),	\
+	MACRO__(0x46D4, ## __VA_ARGS__)
+
+/* RPL-S */
+#define XE_RPLS_IDS(MACRO__, ...)		\
+	MACRO__(0xA780, ## __VA_ARGS__),	\
+	MACRO__(0xA781, ## __VA_ARGS__),	\
+	MACRO__(0xA782, ## __VA_ARGS__),	\
+	MACRO__(0xA783, ## __VA_ARGS__),	\
+	MACRO__(0xA788, ## __VA_ARGS__),	\
+	MACRO__(0xA789, ## __VA_ARGS__),	\
+	MACRO__(0xA78A, ## __VA_ARGS__),	\
+	MACRO__(0xA78B, ## __VA_ARGS__)
+
+/* RPL-U */
+#define XE_RPLU_IDS(MACRO__, ...)		\
+	MACRO__(0xA721, ## __VA_ARGS__),	\
+	MACRO__(0xA7A1, ## __VA_ARGS__),	\
+	MACRO__(0xA7A9, ## __VA_ARGS__),	\
+	MACRO__(0xA7AC, ## __VA_ARGS__),	\
+	MACRO__(0xA7AD, ## __VA_ARGS__)
+
+/* RPL-P */
+#define XE_RPLP_IDS(MACRO__, ...)		\
+	MACRO__(0xA720, ## __VA_ARGS__),	\
+	MACRO__(0xA7A0, ## __VA_ARGS__),	\
+	MACRO__(0xA7A8, ## __VA_ARGS__),	\
+	MACRO__(0xA7AA, ## __VA_ARGS__),	\
+	MACRO__(0xA7AB, ## __VA_ARGS__)
+
+/* DG2 */
+#define XE_DG2_G10_IDS(MACRO__, ...)		\
+	MACRO__(0x5690, ## __VA_ARGS__),	\
+	MACRO__(0x5691, ## __VA_ARGS__),	\
+	MACRO__(0x5692, ## __VA_ARGS__),	\
+	MACRO__(0x56A0, ## __VA_ARGS__),	\
+	MACRO__(0x56A1, ## __VA_ARGS__),	\
+	MACRO__(0x56A2, ## __VA_ARGS__),	\
+	MACRO__(0x56BE, ## __VA_ARGS__),	\
+	MACRO__(0x56BF, ## __VA_ARGS__)
+
+#define XE_DG2_G11_IDS(MACRO__, ...)		\
+	MACRO__(0x5693, ## __VA_ARGS__),	\
+	MACRO__(0x5694, ## __VA_ARGS__),	\
+	MACRO__(0x5695, ## __VA_ARGS__),	\
+	MACRO__(0x56A5, ## __VA_ARGS__),	\
+	MACRO__(0x56A6, ## __VA_ARGS__),	\
+	MACRO__(0x56B0, ## __VA_ARGS__),	\
+	MACRO__(0x56B1, ## __VA_ARGS__),	\
+	MACRO__(0x56BA, ## __VA_ARGS__),	\
+	MACRO__(0x56BB, ## __VA_ARGS__),	\
+	MACRO__(0x56BC, ## __VA_ARGS__),	\
+	MACRO__(0x56BD, ## __VA_ARGS__)
+
+#define XE_DG2_G12_IDS(MACRO__, ...)		\
+	MACRO__(0x5696, ## __VA_ARGS__),	\
+	MACRO__(0x5697, ## __VA_ARGS__),	\
+	MACRO__(0x56A3, ## __VA_ARGS__),	\
+	MACRO__(0x56A4, ## __VA_ARGS__),	\
+	MACRO__(0x56B2, ## __VA_ARGS__),	\
+	MACRO__(0x56B3, ## __VA_ARGS__)
+
+#define XE_DG2_IDS(MACRO__, ...)		\
+	XE_DG2_G10_IDS(MACRO__, ## __VA_ARGS__),\
+	XE_DG2_G11_IDS(MACRO__, ## __VA_ARGS__),\
+	XE_DG2_G12_IDS(MACRO__, ## __VA_ARGS__)
+
+#define XE_ATS_M150_IDS(MACRO__, ...)		\
+	MACRO__(0x56C0, ## __VA_ARGS__),	\
+	MACRO__(0x56C2, ## __VA_ARGS__)
+
+#define XE_ATS_M75_IDS(MACRO__, ...)		\
+	MACRO__(0x56C1, ## __VA_ARGS__)
+
+#define XE_ATS_M_IDS(MACRO__, ...)		\
+	XE_ATS_M150_IDS(MACRO__, ## __VA_ARGS__),\
+	XE_ATS_M75_IDS(MACRO__, ## __VA_ARGS__)
+
+/* ARL */
+#define XE_ARL_IDS(MACRO__, ...)		\
+	MACRO__(0x7D41, ## __VA_ARGS__),	\
+	MACRO__(0x7D51, ## __VA_ARGS__),        \
+	MACRO__(0x7D67, ## __VA_ARGS__),	\
+	MACRO__(0x7DD1, ## __VA_ARGS__),	\
+	MACRO__(0xB640, ## __VA_ARGS__)
+
+/* MTL */
+#define XE_MTL_IDS(MACRO__, ...)		\
+	MACRO__(0x7D40, ## __VA_ARGS__),	\
+	MACRO__(0x7D45, ## __VA_ARGS__),	\
+	MACRO__(0x7D55, ## __VA_ARGS__),	\
+	MACRO__(0x7D60, ## __VA_ARGS__),	\
+	MACRO__(0x7DD5, ## __VA_ARGS__)
+
+/* PVC */
+#define XE_PVC_IDS(MACRO__, ...)		\
+	MACRO__(0x0B69, ## __VA_ARGS__),	\
+	MACRO__(0x0B6E, ## __VA_ARGS__),	\
+	MACRO__(0x0BD4, ## __VA_ARGS__),	\
+	MACRO__(0x0BD5, ## __VA_ARGS__),	\
+	MACRO__(0x0BD6, ## __VA_ARGS__),	\
+	MACRO__(0x0BD7, ## __VA_ARGS__),	\
+	MACRO__(0x0BD8, ## __VA_ARGS__),	\
+	MACRO__(0x0BD9, ## __VA_ARGS__),	\
+	MACRO__(0x0BDA, ## __VA_ARGS__),	\
+	MACRO__(0x0BDB, ## __VA_ARGS__),	\
+	MACRO__(0x0BE0, ## __VA_ARGS__),	\
+	MACRO__(0x0BE1, ## __VA_ARGS__),	\
+	MACRO__(0x0BE5, ## __VA_ARGS__)
+
+#define XE_LNL_IDS(MACRO__, ...) \
+	MACRO__(0x6420, ## __VA_ARGS__), \
+	MACRO__(0x64A0, ## __VA_ARGS__), \
+	MACRO__(0x64B0, ## __VA_ARGS__)
+
+#define XE_BMG_IDS(MACRO__, ...) \
+	MACRO__(0xE202, ## __VA_ARGS__), \
+	MACRO__(0xE20B, ## __VA_ARGS__), \
+	MACRO__(0xE20C, ## __VA_ARGS__), \
+	MACRO__(0xE20D, ## __VA_ARGS__), \
+	MACRO__(0xE212, ## __VA_ARGS__)
+
+#define XE_PTL_IDS(MACRO__, ...) \
+	MACRO__(0xB080, ## __VA_ARGS__), \
+	MACRO__(0xB081, ## __VA_ARGS__), \
+	MACRO__(0xB082, ## __VA_ARGS__), \
+	MACRO__(0xB090, ## __VA_ARGS__), \
+	MACRO__(0xB091, ## __VA_ARGS__), \
+	MACRO__(0xB092, ## __VA_ARGS__), \
+	MACRO__(0xB0A0, ## __VA_ARGS__), \
+	MACRO__(0xB0A1, ## __VA_ARGS__), \
+	MACRO__(0xB0A2, ## __VA_ARGS__), \
+	MACRO__(0xB0B0, ## __VA_ARGS__)
+
+#endif
diff --git a/include/drm/ttm/ttm_bo.h b/include/drm/ttm/ttm_bo.h
index 5804408815be..8ea11cd8df39 100644
--- a/include/drm/ttm/ttm_bo.h
+++ b/include/drm/ttm/ttm_bo.h
@@ -421,6 +421,8 @@ void ttm_bo_unpin(struct ttm_buffer_object *bo);
 int ttm_bo_evict_first(struct ttm_device *bdev,
 		       struct ttm_resource_manager *man,
 		       struct ttm_operation_ctx *ctx);
+int ttm_bo_access(struct ttm_buffer_object *bo, unsigned long offset,
+		  void *buf, int len, int write);
 vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
 			     struct vm_fault *vmf);
 vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
diff --git a/include/uapi/drm/xe_drm.h b/include/uapi/drm/xe_drm.h
index 4a8a4a63e99c..0383b52cbd86 100644
--- a/include/uapi/drm/xe_drm.h
+++ b/include/uapi/drm/xe_drm.h
@@ -1486,6 +1486,7 @@ struct drm_xe_oa_unit {
 	__u64 capabilities;
 #define DRM_XE_OA_CAPS_BASE		(1 << 0)
 #define DRM_XE_OA_CAPS_SYNCS		(1 << 1)
+#define DRM_XE_OA_CAPS_OA_BUFFER_SIZE	(1 << 2)
 
 	/** @oa_timestamp_freq: OA timestamp freq */
 	__u64 oa_timestamp_freq;
@@ -1651,6 +1652,14 @@ enum drm_xe_oa_property_id {
 	 * to the VM bind case.
 	 */
 	DRM_XE_OA_PROPERTY_SYNCS,
+
+	/**
+	 * @DRM_XE_OA_PROPERTY_OA_BUFFER_SIZE: Size of OA buffer to be
+	 * allocated by the driver in bytes. Supported sizes are powers of
+	 * 2 from 128 KiB to 128 MiB. When not specified, a 16 MiB OA
+	 * buffer is allocated by default.
+	 */
+	DRM_XE_OA_PROPERTY_OA_BUFFER_SIZE,
 };
 
 /**