1 files changed, 94 insertions, 63 deletions

diff --git a/drivers/gpu/drm/xe/xe_sync.c b/drivers/gpu/drm/xe/xe_sync.c
index 65f1f1628235..ff74528ca0c6 100644
--- a/drivers/gpu/drm/xe/xe_sync.c
+++ b/drivers/gpu/drm/xe/xe_sync.c
@@ -12,9 +12,9 @@
 
 #include <drm/drm_print.h>
 #include <drm/drm_syncobj.h>
-#include <drm/xe_drm.h>
+#include <uapi/drm/xe_drm.h>
 
-#include "xe_device_types.h"
+#include "xe_device.h"
 #include "xe_exec_queue.h"
 #include "xe_macros.h"
 #include "xe_sched_job_types.h"
@@ -53,14 +53,19 @@ static struct xe_user_fence *user_fence_create(struct xe_device *xe, u64 addr,
 					       u64 value)
 {
 	struct xe_user_fence *ufence;
+	u64 __user *ptr = u64_to_user_ptr(addr);
+	u64 __maybe_unused prefetch_val;
 
-	ufence = kmalloc(sizeof(*ufence), GFP_KERNEL);
+	if (get_user(prefetch_val, ptr))
+		return ERR_PTR(-EFAULT);
+
+	ufence = kzalloc(sizeof(*ufence), GFP_KERNEL);
 	if (!ufence)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	ufence->xe = xe;
 	kref_init(&ufence->refcount);
-	ufence->addr = u64_to_user_ptr(addr);
+	ufence->addr = ptr;
 	ufence->value = value;
 	ufence->mm = current->mm;
 	mmgrab(ufence->mm);
@@ -72,16 +77,22 @@ static void user_fence_worker(struct work_struct *w)
 {
 	struct xe_user_fence *ufence = container_of(w, struct xe_user_fence, worker);
 
+	WRITE_ONCE(ufence->signalled, 1);
 	if (mmget_not_zero(ufence->mm)) {
 		kthread_use_mm(ufence->mm);
 		if (copy_to_user(ufence->addr, &ufence->value, sizeof(ufence->value)))
 			XE_WARN_ON("Copy to user failed");
 		kthread_unuse_mm(ufence->mm);
 		mmput(ufence->mm);
+	} else {
+		drm_dbg(&ufence->xe->drm, "mmget_not_zero() failed, ufence wasn't signaled\n");
 	}
 
+	/*
+	 * Wake up waiters only after updating the ufence state, allowing the UMD
+	 * to safely reuse the same ufence without encountering -EBUSY errors.
+	 */
 	wake_up_all(&ufence->xe->ufence_wq);
-	WRITE_ONCE(ufence->signalled, 1);
 	user_fence_put(ufence);
 }
 
@@ -102,6 +113,8 @@ static void user_fence_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
 int xe_sync_entry_parse(struct xe_device *xe, struct xe_file *xef,
 			struct xe_sync_entry *sync,
 			struct drm_xe_sync __user *sync_user,
+			struct drm_syncobj *ufence_syncobj,
+			u64 ufence_timeline_value,
 			unsigned int flags)
 {
 	struct drm_xe_sync sync_in;
@@ -181,10 +194,15 @@ int xe_sync_entry_parse(struct xe_device *xe, struct xe_file *xef,
 		if (exec) {
 			sync->addr = sync_in.addr;
 		} else {
+			sync->ufence_timeline_value = ufence_timeline_value;
 			sync->ufence = user_fence_create(xe, sync_in.addr,
 							 sync_in.timeline_value);
-			if (XE_IOCTL_DBG(xe, !sync->ufence))
+			if (XE_IOCTL_DBG(xe, IS_ERR(sync->ufence)))
+				return PTR_ERR(sync->ufence);
+			sync->ufence_chain_fence = dma_fence_chain_alloc();
+			if (!sync->ufence_chain_fence)
 				return -ENOMEM;
+			sync->ufence_syncobj = ufence_syncobj;
 		}
 
 		break;
@@ -199,27 +217,13 @@ int xe_sync_entry_parse(struct xe_device *xe, struct xe_file *xef,
 
 	return 0;
 }
-
-int xe_sync_entry_wait(struct xe_sync_entry *sync)
-{
-	if (sync->fence)
-		dma_fence_wait(sync->fence, true);
-
-	return 0;
-}
+ALLOW_ERROR_INJECTION(xe_sync_entry_parse, ERRNO);
 
 int xe_sync_entry_add_deps(struct xe_sync_entry *sync, struct xe_sched_job *job)
 {
-	int err;
-
-	if (sync->fence) {
-		err = drm_sched_job_add_dependency(&job->drm,
-						   dma_fence_get(sync->fence));
-		if (err) {
-			dma_fence_put(sync->fence);
-			return err;
-		}
-	}
+	if (sync->fence)
+		return  drm_sched_job_add_dependency(&job->drm,
+						     dma_fence_get(sync->fence));
 
 	return 0;
 }
@@ -242,7 +246,12 @@ void xe_sync_entry_signal(struct xe_sync_entry *sync, struct dma_fence *fence)
 	} else if (sync->ufence) {
 		int err;
 
-		dma_fence_get(fence);
+		drm_syncobj_add_point(sync->ufence_syncobj,
+				      sync->ufence_chain_fence,
+				      fence, sync->ufence_timeline_value);
+		sync->ufence_chain_fence = NULL;
+
+		fence = drm_syncobj_fence_get(sync->ufence_syncobj);
 		user_fence_get(sync->ufence);
 		err = dma_fence_add_callback(fence, &sync->ufence->cb,
 					     user_fence_cb);
@@ -260,11 +269,10 @@ void xe_sync_entry_cleanup(struct xe_sync_entry *sync)
 {
 	if (sync->syncobj)
 		drm_syncobj_put(sync->syncobj);
-	if (sync->fence)
-		dma_fence_put(sync->fence);
-	if (sync->chain_fence)
-		dma_fence_put(&sync->chain_fence->base);
-	if (sync->ufence)
+	dma_fence_put(sync->fence);
+	dma_fence_chain_free(sync->chain_fence);
+	dma_fence_chain_free(sync->ufence_chain_fence);
+	if (!IS_ERR_OR_NULL(sync->ufence))
 		user_fence_put(sync->ufence);
 }
 
@@ -289,56 +297,79 @@ xe_sync_in_fence_get(struct xe_sync_entry *sync, int num_sync,
 	struct dma_fence **fences = NULL;
 	struct dma_fence_array *cf = NULL;
 	struct dma_fence *fence;
-	int i, num_in_fence = 0, current_fence = 0;
+	int i, num_fence = 0, current_fence = 0;
 
 	lockdep_assert_held(&vm->lock);
 
-	/* Count in-fences */
-	for (i = 0; i < num_sync; ++i) {
-		if (sync[i].fence) {
-			++num_in_fence;
-			fence = sync[i].fence;
+	/* Reject in fences */
+	for (i = 0; i < num_sync; ++i)
+		if (sync[i].fence)
+			return ERR_PTR(-EOPNOTSUPP);
+
+	if (q->flags & EXEC_QUEUE_FLAG_VM) {
+		struct xe_exec_queue *__q;
+		struct xe_tile *tile;
+		u8 id;
+
+		for_each_tile(tile, vm->xe, id)
+			num_fence += (1 + XE_MAX_GT_PER_TILE);
+
+		fences = kmalloc_array(num_fence, sizeof(*fences),
+				       GFP_KERNEL);
+		if (!fences)
+			return ERR_PTR(-ENOMEM);
+
+		fences[current_fence++] =
+			xe_exec_queue_last_fence_get(q, vm);
+		for_each_tlb_inval(i)
+			fences[current_fence++] =
+				xe_exec_queue_tlb_inval_last_fence_get(q, vm, i);
+		list_for_each_entry(__q, &q->multi_gt_list,
+				    multi_gt_link) {
+			fences[current_fence++] =
+				xe_exec_queue_last_fence_get(__q, vm);
+			for_each_tlb_inval(i)
+				fences[current_fence++] =
+					xe_exec_queue_tlb_inval_last_fence_get(__q, vm, i);
 		}
-	}
 
-	/* Easy case... */
-	if (!num_in_fence) {
-		fence = xe_exec_queue_last_fence_get(q, vm);
-		return fence;
-	}
+		xe_assert(vm->xe, current_fence == num_fence);
+		cf = dma_fence_array_create(num_fence, fences,
+					    dma_fence_context_alloc(1),
+					    1, false);
+		if (!cf)
+			goto err_out;
 
-	/* Create composite fence */
-	fences = kmalloc_array(num_in_fence + 1, sizeof(*fences), GFP_KERNEL);
-	if (!fences)
-		return ERR_PTR(-ENOMEM);
-	for (i = 0; i < num_sync; ++i) {
-		if (sync[i].fence) {
-			dma_fence_get(sync[i].fence);
-			fences[current_fence++] = sync[i].fence;
-		}
-	}
-	fences[current_fence++] = xe_exec_queue_last_fence_get(q, vm);
-	cf = dma_fence_array_create(num_in_fence, fences,
-				    vm->composite_fence_ctx,
-				    vm->composite_fence_seqno++,
-				    false);
-	if (!cf) {
-		--vm->composite_fence_seqno;
-		goto err_out;
+		return &cf->base;
 	}
 
-	return &cf->base;
+	fence = xe_exec_queue_last_fence_get(q, vm);
+	return fence;
 
 err_out:
 	while (current_fence)
 		dma_fence_put(fences[--current_fence]);
 	kfree(fences);
-	kfree(cf);
 
 	return ERR_PTR(-ENOMEM);
 }
 
 /**
+ * __xe_sync_ufence_get() - Get user fence from user fence
+ * @ufence: input user fence
+ *
+ * Get a user fence reference from user fence
+ *
+ * Return: xe_user_fence pointer with reference
+ */
+struct xe_user_fence *__xe_sync_ufence_get(struct xe_user_fence *ufence)
+{
+	user_fence_get(ufence);
+
+	return ufence;
+}
+
+/**
  * xe_sync_ufence_get() - Get user fence from sync
  * @sync: input sync
  *