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 <[email protected]>

# -----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 <[email protected]>" [unknown]
# gpg:                 aka "Rodrigo Vivi <[email protected]>" [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 <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

Author: airlied

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

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:

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)

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)

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 = {

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 \

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

drivers/gpu/drm/xe/abi/guc_actions_sriov_abi.h

@@ -501,6 +501,44 @@
 #define VF2GUC_VF_RESET_RESPONSE_MSG_LEN		GUC_HXG_RESPONSE_MSG_MIN_LEN
 #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
  *

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:
  */

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)