[PATCH v4 4/5] drm/xe: Don't allow in fences on zero batch exec or zero binds

[Matthew Brost <matthew.brost@intel.com>]

The zero batch or zero binds mechanism was added to implement queue
idling in Mesa. In this case, output fences are attached, which can be
waited upon to indicate that the queue is idle once they signal.

As part of this, we added the ability to install input fences on zero
batch execs or zero binds for interface congruence. However, upon
inspection, this implementation doesn't work correctly, as multiple
composite fences could be chained together, which is disallowed.

While this could be fixed, it would be rather complicated. Since the use
case for input fences on zero batch execs or zero binds isn't actually
used, it's better to remove support and disallow it.

Signed-off-by: Matthew Brost <matthew.brost@intel.com>
---
 drivers/gpu/drm/xe/xe_sync.c | 101 +++++++++++++----------------------
 1 file changed, 36 insertions(+), 65 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_sync.c b/drivers/gpu/drm/xe/xe_sync.c
index 70706362c11a..f50134bedf8a 100644
--- a/drivers/gpu/drm/xe/xe_sync.c
+++ b/drivers/gpu/drm/xe/xe_sync.c
@@ -288,84 +288,55 @@ xe_sync_in_fence_get(struct xe_sync_entry *sync, int num_sync,
 
 	lockdep_assert_held(&vm->lock);
 
-	/* Count in-fences */
-	for (i = 0; i < num_sync; ++i) {
-		if (sync[i].fence) {
-			++num_fence;
-			fence = sync[i].fence;
-		}
-	}
-
-	/* Easy case... */
-	if (!num_fence) {
-		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);
-
+	/* 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);
+				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);
-			}
-
-			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;
-
-			return &cf->base;
+					xe_exec_queue_tlb_inval_last_fence_get(__q, vm, i);
 		}
 
-		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 - FIXME - the below code doesn't work. This is
-	 * unused in Mesa so we are ok for the moment. Perhaps we just disable
-	 * this entire code path if number of in fences != 0.
-	 */
-	fences = kmalloc_array(num_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;
-		}
+		return &cf->base;
 	}
-	fences[current_fence++] = xe_exec_queue_last_fence_get(q, vm);
-	cf = dma_fence_array_create(num_fence, fences,
-				    dma_fence_context_alloc(1), 1, false);
-	if (!cf)
-		goto err_out;
 
-	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);
 }
-- 
2.34.1