[PATCH 2/4] drm/sched: add drm_sched_prealloc_dependency_slots

[Christian König]

Sometimes drivers need to be able to submit multiple jobs which depend on
each other to different schedulers at the same time, but using
drm_sched_job_add_dependency() can't fail any more after the first job is
initialized.

This function preallocate memory for dependency slots so that no ENOMEM
can come later while adding dependencies.

Signed-off-by: Christian König <christian.koenig at amd.com>
---
 drivers/gpu/drm/scheduler/sched_main.c | 42 +++++++++++++++++++++++++-
 include/drm/gpu_scheduler.h            |  2 ++
 2 files changed, 43 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c
index cf200b1b643e..37082f52f43f 100644
--- a/drivers/gpu/drm/scheduler/sched_main.c
+++ b/drivers/gpu/drm/scheduler/sched_main.c
@@ -858,6 +858,43 @@ void drm_sched_job_arm(struct drm_sched_job *job)
 }
 EXPORT_SYMBOL(drm_sched_job_arm);
 
+/**
+ * drm_sched_job_prealloc_dependency_slots - avoid ENOMEM on adding dependencies
+ * @job: scheduler job where dependencies will be added
+ * @num_slots: number of slots to reserve
+  *
+ * Sometimes drivers need to be able to submit multiple jobs which depend on
+ * each other to different schedulers at the same time, but using
+ * drm_sched_job_add_dependency() can't fail any more after the first job is
+ * initialized.
+ *
+ * This function preallocate memory for dependency slots so that no ENOMEM can
+ * come later while adding dependencies.
+ *
+ * Return:
+ * 0 on success, or an error on failing to expand the array.
+ */
+int drm_sched_job_prealloc_dependency_slots(struct drm_sched_job *job,
+					    unsigned int num_slots)
+{
+	int ret;
+	u32 id;
+
+	/*
+	 * This works because NULL entries are not returned by xa_for_each. So
+	 * drm_sched_job_add_dependency() will ignore those while checking for
+	 * duplicates, but can then use the entry when storing the new fence.
+	 */
+	while (num_slots--) {
+		ret = xa_alloc(&job->dependencies, &id, NULL, xa_limit_32b,
+			       GFP_KERNEL);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
+EXPORT_SYMBOL(drm_sched_job_prealloc_dependency_slots);
+
 /**
  * drm_sched_job_add_dependency - adds the fence as a job dependency
  * @job: scheduler job to add the dependencies to
@@ -883,6 +920,9 @@ int drm_sched_job_add_dependency(struct drm_sched_job *job,
 	 */
 	xa_lock(&job->dependencies);
 	xas_for_each(&xas, entry, ULONG_MAX) {
+		if (xa_is_zero(entry))
+		    break;
+
 		if (entry->context != fence->context)
 			continue;
 
@@ -909,7 +949,7 @@ int drm_sched_job_add_dependency(struct drm_sched_job *job,
 	if (xas_error(&xas))
 		dma_fence_put(fence);
 	else
-		WARN_ON(entry);
+		WARN_ON(entry && !xa_is_zero(entry));
 
 	return xas_error(&xas);
 }
diff --git a/include/drm/gpu_scheduler.h b/include/drm/gpu_scheduler.h
index d860db087ea5..a560a00c6275 100644
--- a/include/drm/gpu_scheduler.h
+++ b/include/drm/gpu_scheduler.h
@@ -632,6 +632,8 @@ int drm_sched_job_init(struct drm_sched_job *job,
 		       u32 credits, void *owner);
 void drm_sched_job_arm(struct drm_sched_job *job);
 void drm_sched_entity_push_job(struct drm_sched_job *sched_job);
+int drm_sched_job_prealloc_dependency_slots(struct drm_sched_job *job,
+					    unsigned int num_slots);
 int drm_sched_job_add_dependency(struct drm_sched_job *job,
 				 struct dma_fence *fence);
 int drm_sched_job_add_syncobj_dependency(struct drm_sched_job *job,
-- 
2.34.1