Merge tag 'drm-xe-next-fixes-2026-06-17' of https://gitlab.freedesktop.org/drm/xe/kernel into drm-next

- Set TTM beneficial order to 9 in Xe
- Several error path cleanups
- Fix TDR for unstarted jobs on kernel queues
- Several TLB invalidation fixes related to suspending LR queues
- Some small RAS fixes
- Multi-queue suspend fix for LR queues
- Revert inclusion of NVL_S firmware

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Matthew Brost <matthew.brost@intel.com>
Link: https://patch.msgid.link/ajLy2brwvOZEFNNN@gsse-cloud1.jf.intel.com

Author: airlied

drivers/gpu/drm/xe/Makefile

@@ -16,14 +16,14 @@ subdir-ccflags-y += -I$(obj) -I$(src)
 hostprogs := xe_gen_wa_oob
 generated_oob := $(obj)/generated/xe_wa_oob.c $(obj)/generated/xe_wa_oob.h
 quiet_cmd_wa_oob = GEN     $(notdir $(generated_oob))
-      cmd_wa_oob = mkdir -p $(@D); $^ $(generated_oob)
+      cmd_wa_oob = mkdir -p $(@D); $(obj)/xe_gen_wa_oob $(src)/xe_wa_oob.rules $(generated_oob)
 $(obj)/generated/%_wa_oob.c $(obj)/generated/%_wa_oob.h: $(obj)/xe_gen_wa_oob \
 		 $(src)/xe_wa_oob.rules
 	$(call cmd,wa_oob)
 
 generated_device_oob := $(obj)/generated/xe_device_wa_oob.c $(obj)/generated/xe_device_wa_oob.h
 quiet_cmd_device_wa_oob = GEN	$(notdir $(generated_device_oob))
-      cmd_device_wa_oob = mkdir -p $(@D); $^ $(generated_device_oob)
+      cmd_device_wa_oob = mkdir -p $(@D); $(obj)/xe_gen_wa_oob $(src)/xe_device_wa_oob.rules $(generated_device_oob)
 $(obj)/generated/%_device_wa_oob.c $(obj)/generated/%_device_wa_oob.h: $(obj)/xe_gen_wa_oob \
 		 $(src)/xe_device_wa_oob.rules
 	$(call cmd,device_wa_oob)

drivers/gpu/drm/xe/regs/xe_gtt_defs.h

@@ -9,7 +9,11 @@
 #define XELPG_GGTT_PTE_PAT0	BIT_ULL(52)
 #define XELPG_GGTT_PTE_PAT1	BIT_ULL(53)
 
-#define XE_PTE_ADDR_MASK	GENMASK_ULL(51, 12)
+/*
+ * Mask for PTE address bits [51:shift].
+ * shift is the lower address boundary of page.
+ */
+#define XE_PAGE_ADDR_MASK(shift)	GENMASK_ULL(51, (shift))
 #define GGTT_PTE_VFID		GENMASK_ULL(11, 2)
 
 #define GUC_GGTT_TOP		0xFEE00000

drivers/gpu/drm/xe/xe_device.c

@@ -526,7 +526,8 @@ int xe_device_init_early(struct xe_device *xe)
 
 	err = ttm_device_init(&xe->ttm, &xe_ttm_funcs, xe->drm.dev,
 			      xe->drm.anon_inode->i_mapping,
-			      xe->drm.vma_offset_manager, 0);
+			      xe->drm.vma_offset_manager,
+			      TTM_ALLOCATION_POOL_BENEFICIAL_ORDER(get_order(SZ_2M)));
 	if (err)
 		return err;
 

drivers/gpu/drm/xe/xe_drm_ras.c

@@ -80,7 +80,7 @@ static struct xe_drm_ras_counter *allocate_and_copy_counters(struct xe_device *x
 	struct xe_drm_ras_counter *counter;
 	int i;
 
-	counter = kcalloc(DRM_XE_RAS_ERR_COMP_MAX, sizeof(*counter), GFP_KERNEL);
+	counter = drmm_kcalloc(&xe->drm, DRM_XE_RAS_ERR_COMP_MAX, sizeof(*counter), GFP_KERNEL);
 	if (!counter)
 		return ERR_PTR(-ENOMEM);
 
@@ -131,54 +131,47 @@ static int assign_node_params(struct xe_device *xe, struct drm_ras_node *node,
 	return 0;
 }
 
-static void cleanup_node_param(struct xe_drm_ras *ras, const enum drm_xe_ras_error_severity severity)
+static void cleanup_node_param(struct drm_ras_node *node)
 {
-	struct drm_ras_node *node = &ras->node[severity];
-
-	kfree(ras->info[severity]);
-	ras->info[severity] = NULL;
-
 	kfree(node->device_name);
 	node->device_name = NULL;
 }
 
+static void cleanup_node(struct drm_device *drm, void *node)
+{
+	drm_ras_node_unregister(node);
+	cleanup_node_param(node);
+}
+
 static int register_nodes(struct xe_device *xe)
 {
 	struct xe_drm_ras *ras = &xe->ras;
-	int i;
+	struct drm_ras_node *node;
+	int i, ret;
 
 	for_each_error_severity(i) {
-		struct drm_ras_node *node = &ras->node[i];
-		int ret;
+		node = &ras->node[i];
 
 		ret = assign_node_params(xe, node, i);
-		if (ret) {
-			cleanup_node_param(ras, i);
-			return ret;
-		}
+		if (ret)
+			goto free_param;
 
 		ret = drm_ras_node_register(node);
-		if (ret) {
-			cleanup_node_param(ras, i);
-			return ret;
-		}
+		if (ret)
+			goto free_param;
+
+		ret = drmm_add_action_or_reset(&xe->drm, cleanup_node, node);
+		if (ret)
+			goto null_info;
 	}
 
 	return 0;
-}
-
-static void xe_drm_ras_unregister_nodes(struct drm_device *device, void *arg)
-{
-	struct xe_device *xe = arg;
-	struct xe_drm_ras *ras = &xe->ras;
-	int i;
-
-	for_each_error_severity(i) {
-		struct drm_ras_node *node = &ras->node[i];
 
-		drm_ras_node_unregister(node);
-		cleanup_node_param(ras, i);
-	}
+free_param:
+	cleanup_node_param(node);
+null_info:
+	ras->info[i] = NULL;
+	return ret;
 }
 
 /**
@@ -207,11 +200,5 @@ int xe_drm_ras_init(struct xe_device *xe)
 		return err;
 	}
 
-	err = drmm_add_action_or_reset(&xe->drm, xe_drm_ras_unregister_nodes, xe);
-	if (err) {
-		drm_err(&xe->drm, "Failed to add action for Xe DRM RAS (%pe)\n", ERR_PTR(err));
-		return err;
-	}
-
 	return 0;
 }

drivers/gpu/drm/xe/xe_exec_queue.h

@@ -162,21 +162,4 @@ int xe_exec_queue_contexts_hwsp_rebase(struct xe_exec_queue *q, void *scratch);
 struct xe_lrc *xe_exec_queue_lrc(struct xe_exec_queue *q);
 struct xe_lrc *xe_exec_queue_get_lrc(struct xe_exec_queue *q, u16 idx);
 
-/**
- * xe_exec_queue_idle_skip_suspend() - Can exec queue skip suspend
- * @q: The exec_queue
- *
- * If an exec queue is not parallel and is idle, the suspend steps can be
- * skipped in the submission backend immediatley signaling the suspend fence.
- * Parallel queues cannot skip this step due to limitations in the submission
- * backend.
- *
- * Return: True if exec queue is idle and can skip suspend steps, False
- * otherwise
- */
-static inline bool xe_exec_queue_idle_skip_suspend(struct xe_exec_queue *q)
-{
-	return !xe_exec_queue_is_parallel(q) && xe_exec_queue_is_idle(q);
-}
-
 #endif

drivers/gpu/drm/xe/xe_guc_capture.c

@@ -461,8 +461,14 @@ static void guc_capture_alloc_steered_lists(struct xe_guc *guc)
 	if (!list || guc->capture->extlists)
 		return;
 
-	total = bitmap_weight(gt->fuse_topo.g_dss_mask, sizeof(gt->fuse_topo.g_dss_mask) * 8) *
-		guc_capture_get_steer_reg_num(guc_to_xe(guc));
+	{
+		xe_dss_mask_t all_dss;
+
+		total = bitmap_weighted_or(all_dss, gt->fuse_topo.g_dss_mask,
+					   gt->fuse_topo.c_dss_mask,
+					   XE_MAX_DSS_FUSE_BITS) *
+			guc_capture_get_steer_reg_num(guc_to_xe(guc));
+	}
 
 	if (!total)
 		return;

drivers/gpu/drm/xe/xe_guc_submit.c

@@ -71,7 +71,6 @@ exec_queue_to_guc(struct xe_exec_queue *q)
 #define EXEC_QUEUE_STATE_WEDGED			(1 << 8)
 #define EXEC_QUEUE_STATE_BANNED			(1 << 9)
 #define EXEC_QUEUE_STATE_PENDING_RESUME		(1 << 10)
-#define EXEC_QUEUE_STATE_IDLE_SKIP_SUSPEND	(1 << 11)
 
 static bool exec_queue_registered(struct xe_exec_queue *q)
 {
@@ -158,6 +157,11 @@ static void set_exec_queue_banned(struct xe_exec_queue *q)
 	atomic_or(EXEC_QUEUE_STATE_BANNED, &q->guc->state);
 }
 
+static void clear_exec_queue_banned(struct xe_exec_queue *q)
+{
+	atomic_andnot(EXEC_QUEUE_STATE_BANNED, &q->guc->state);
+}
+
 static bool exec_queue_suspended(struct xe_exec_queue *q)
 {
 	return atomic_read(&q->guc->state) & EXEC_QUEUE_STATE_SUSPENDED;
@@ -218,21 +222,6 @@ static void clear_exec_queue_pending_resume(struct xe_exec_queue *q)
 	atomic_and(~EXEC_QUEUE_STATE_PENDING_RESUME, &q->guc->state);
 }
 
-static bool exec_queue_idle_skip_suspend(struct xe_exec_queue *q)
-{
-	return atomic_read(&q->guc->state) & EXEC_QUEUE_STATE_IDLE_SKIP_SUSPEND;
-}
-
-static void set_exec_queue_idle_skip_suspend(struct xe_exec_queue *q)
-{
-	atomic_or(EXEC_QUEUE_STATE_IDLE_SKIP_SUSPEND, &q->guc->state);
-}
-
-static void clear_exec_queue_idle_skip_suspend(struct xe_exec_queue *q)
-{
-	atomic_and(~EXEC_QUEUE_STATE_IDLE_SKIP_SUSPEND, &q->guc->state);
-}
-
 static bool exec_queue_killed_or_banned_or_wedged(struct xe_exec_queue *q)
 {
 	return (atomic_read(&q->guc->state) &
@@ -1157,7 +1146,7 @@ static void submit_exec_queue(struct xe_exec_queue *q, struct xe_sched_job *job)
 	if (!job->restore_replay || job->last_replay) {
 		if (xe_exec_queue_is_parallel(q))
 			wq_item_append(q);
-		else if (!exec_queue_idle_skip_suspend(q))
+		else
 			xe_lrc_set_ring_tail(lrc, lrc->ring.tail);
 		job->last_replay = false;
 	}
@@ -1167,9 +1156,12 @@ static void submit_exec_queue(struct xe_exec_queue *q, struct xe_sched_job *job)
 
 	/*
 	 * All queues in a multi-queue group will use the primary queue
-	 * of the group to interface with GuC.
+	 * of the group to interface with GuC. If primay is suspended,
+	 * just return. Jobs will get scheduled once primary is resumed.
 	 */
 	q = xe_exec_queue_multi_queue_primary(q);
+	if (exec_queue_suspended(q))
+		return;
 
 	if (!exec_queue_enabled(q) && !exec_queue_suspended(q)) {
 		action[len++] = XE_GUC_ACTION_SCHED_CONTEXT_MODE_SET;
@@ -1376,7 +1368,8 @@ static bool check_timeout(struct xe_exec_queue *q, struct xe_sched_job *job)
 			   xe_sched_job_seqno(job), xe_sched_job_lrc_seqno(job),
 			   q->guc->id);
 
-		return xe_sched_invalidate_job(job, 2);
+		/* GuC never scheduled this job - let the caller trigger a GT reset. */
+		return true;
 	}
 
 	ctx_timestamp = lower_32_bits(xe_lrc_timestamp(q->lrc[0]));
@@ -1473,6 +1466,21 @@ static void disable_scheduling(struct xe_exec_queue *q, bool immediate)
 			       G2H_LEN_DW_SCHED_CONTEXT_MODE_SET, 1);
 }
 
+/*
+ * Recover via GT reset for a kernel queue, or for a GuC scheduling failure (job
+ * never started) on a queue that was not already killed or banned. An already
+ * banned queue must stay banned, so its unstarted jobs do not clear the ban or
+ * trigger a reset.
+ */
+static bool timeout_needs_gt_reset(struct xe_exec_queue *q, struct xe_sched_job *job,
+				   bool skip_timeout_check)
+{
+	if (q->flags & EXEC_QUEUE_FLAG_KERNEL)
+		return true;
+
+	return !skip_timeout_check && !xe_sched_job_started(job);
+}
+
 static enum drm_gpu_sched_stat
 guc_exec_queue_timedout_job(struct drm_sched_job *drm_job)
 {
@@ -1621,19 +1629,19 @@ guc_exec_queue_timedout_job(struct drm_sched_job *drm_job)
 			       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
-	 * somethings has gone wrong and the GT needs a reset
-	 */
-	xe_gt_WARN(q->gt, q->flags & EXEC_QUEUE_FLAG_KERNEL,
-		   "Kernel-submitted job timed out\n");
-	xe_gt_WARN(q->gt, q->flags & EXEC_QUEUE_FLAG_VM && !exec_queue_killed(q),
-		   "VM job timed out on non-killed execqueue\n");
-	if (!wedged && (q->flags & EXEC_QUEUE_FLAG_KERNEL ||
-			(q->flags & EXEC_QUEUE_FLAG_VM && !exec_queue_killed(q)))) {
-		if (!xe_sched_invalidate_job(job, 2)) {
-			xe_gt_reset_async(q->gt);
-			goto rearm;
+	if (!wedged) {
+		if (timeout_needs_gt_reset(q, job, skip_timeout_check)) {
+			if (!xe_sched_invalidate_job(job, 2)) {
+				clear_exec_queue_banned(q);
+				xe_gt_reset_async(q->gt);
+				goto rearm;
+			}
+			if (q->flags & EXEC_QUEUE_FLAG_KERNEL) {
+				xe_gt_WARN(q->gt, true, "Kernel-submitted job timed out\n");
+				xe_device_declare_wedged(gt_to_xe(q->gt));
+			}
+		} else if (q->flags & EXEC_QUEUE_FLAG_VM && !exec_queue_killed(q)) {
+			xe_gt_WARN(q->gt, true, "VM job timed out on non-killed execqueue\n");
 		}
 	}
 
@@ -1812,10 +1820,9 @@ static void __guc_exec_queue_process_msg_suspend(struct xe_sched_msg *msg)
 {
 	struct xe_exec_queue *q = msg->private_data;
 	struct xe_guc *guc = exec_queue_to_guc(q);
-	bool idle_skip_suspend = xe_exec_queue_idle_skip_suspend(q);
 
-	if (!idle_skip_suspend && guc_exec_queue_allowed_to_change_state(q) &&
-	    !exec_queue_suspended(q) && exec_queue_enabled(q)) {
+	if (guc_exec_queue_allowed_to_change_state(q) && !exec_queue_suspended(q) &&
+	    exec_queue_enabled(q)) {
 		wait_event(guc->ct.wq, vf_recovery(guc) ||
 			   ((q->guc->resume_time != RESUME_PENDING ||
 			   xe_guc_read_stopped(guc)) && !exec_queue_pending_disable(q)));
@@ -1834,56 +1841,24 @@ static void __guc_exec_queue_process_msg_suspend(struct xe_sched_msg *msg)
 			disable_scheduling(q, false);
 		}
 	} else if (q->guc->suspend_pending) {
-		if (idle_skip_suspend)
-			set_exec_queue_idle_skip_suspend(q);
 		set_exec_queue_suspended(q);
 		suspend_fence_signal(q);
 	}
 }
 
-static void sched_context(struct xe_exec_queue *q)
-{
-	struct xe_guc *guc = exec_queue_to_guc(q);
-	struct xe_lrc *lrc = q->lrc[0];
-	u32 action[] = {
-		XE_GUC_ACTION_SCHED_CONTEXT,
-		q->guc->id,
-	};
-
-	xe_gt_assert(guc_to_gt(guc), !xe_exec_queue_is_parallel(q));
-	xe_gt_assert(guc_to_gt(guc), !exec_queue_destroyed(q));
-	xe_gt_assert(guc_to_gt(guc), exec_queue_registered(q));
-	xe_gt_assert(guc_to_gt(guc), !exec_queue_pending_disable(q));
-
-	trace_xe_exec_queue_submit(q);
-
-	xe_lrc_set_ring_tail(lrc, lrc->ring.tail);
-	xe_guc_ct_send(&guc->ct, action, ARRAY_SIZE(action), 0, 0);
-}
-
 static void __guc_exec_queue_process_msg_resume(struct xe_sched_msg *msg)
 {
 	struct xe_exec_queue *q = msg->private_data;
 
 	if (guc_exec_queue_allowed_to_change_state(q)) {
 		clear_exec_queue_suspended(q);
 		if (!exec_queue_enabled(q)) {
-			if (exec_queue_idle_skip_suspend(q)) {
-				struct xe_lrc *lrc = q->lrc[0];
-
-				clear_exec_queue_idle_skip_suspend(q);
-				xe_lrc_set_ring_tail(lrc, lrc->ring.tail);
-			}
 			q->guc->resume_time = RESUME_PENDING;
 			set_exec_queue_pending_resume(q);
 			enable_scheduling(q);
-		} else if (exec_queue_idle_skip_suspend(q)) {
-			clear_exec_queue_idle_skip_suspend(q);
-			sched_context(q);
 		}
 	} else {
 		clear_exec_queue_suspended(q);
-		clear_exec_queue_idle_skip_suspend(q);
 	}
 }
 
@@ -2840,8 +2815,8 @@ static void handle_sched_done(struct xe_guc *guc, struct xe_exec_queue *q,
 		xe_gt_assert(guc_to_gt(guc), exec_queue_pending_disable(q));
 
 		if (q->guc->suspend_pending) {
-			suspend_fence_signal(q);
 			clear_exec_queue_pending_disable(q);
+			suspend_fence_signal(q);
 		} else {
 			if (exec_queue_banned(q)) {
 				smp_wmb();