[PATCH 09/12] drm/i915: Create a kmem_cache to allocate struct i915_priolist from

[Chris Wilson]

The i915_priolist are allocated within an atomic context on a path where
we wish to minimise latency. If we use a dedicated kmem_cache, we have
the advantage of a local freelist from which to service new requests
that should keep the latency impact of an allocation small. Though
currently we expect the majority of requests to be at default priority
(and so hit the preallocate priolist), once userspace starts using
priorities they are likely to use many fine grained policies improving
the utilisation of a private slab.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h                  | 1 +
 drivers/gpu/drm/i915/i915_gem.c                  | 9 ++++++++-
 drivers/gpu/drm/i915/i915_guc_submission.c       | 2 +-
 drivers/gpu/drm/i915/intel_lrc.c                 | 4 ++--
 drivers/gpu/drm/i915/selftests/mock_gem_device.c | 9 ++++++++-
 5 files changed, 20 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index a6f20471b4cd..08ee5c8834fb 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2027,6 +2027,7 @@ struct drm_i915_private {
 	struct kmem_cache *vmas;
 	struct kmem_cache *requests;
 	struct kmem_cache *dependencies;
+	struct kmem_cache *priorities;
 
 	const struct intel_device_info info;
 
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 07db08cc6be0..02adf8241394 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -4866,12 +4866,16 @@ i915_gem_load_init(struct drm_i915_private *dev_priv)
 	if (!dev_priv->dependencies)
 		goto err_requests;
 
+	dev_priv->priorities = KMEM_CACHE(i915_priolist, SLAB_HWCACHE_ALIGN);
+	if (!dev_priv->priorities)
+		goto err_dependencies;
+
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
 	err = i915_gem_timeline_init__global(dev_priv);
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 	if (err)
-		goto err_dependencies;
+		goto err_priorities;
 
 	INIT_LIST_HEAD(&dev_priv->context_list);
 	INIT_WORK(&dev_priv->mm.free_work, __i915_gem_free_work);
@@ -4895,6 +4899,8 @@ i915_gem_load_init(struct drm_i915_private *dev_priv)
 
 	return 0;
 
+err_priorities:
+	kmem_cache_destroy(dev_priv->priorities);
 err_dependencies:
 	kmem_cache_destroy(dev_priv->dependencies);
 err_requests:
@@ -4918,6 +4924,7 @@ void i915_gem_load_cleanup(struct drm_i915_private *dev_priv)
 	WARN_ON(!list_empty(&dev_priv->gt.timelines));
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
+	kmem_cache_destroy(dev_priv->priorities);
 	kmem_cache_destroy(dev_priv->dependencies);
 	kmem_cache_destroy(dev_priv->requests);
 	kmem_cache_destroy(dev_priv->vmas);
diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index 3b9cdb0907c2..b3da056ea8f1 100644
--- a/drivers/gpu/drm/i915/i915_guc_submission.c
+++ b/drivers/gpu/drm/i915/i915_guc_submission.c
@@ -704,7 +704,7 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine)
 		rb_erase(&p->node, &engine->execlist_queue);
 		INIT_LIST_HEAD(&p->requests);
 		if (p->priority != I915_PRIORITY_NORMAL)
-			kfree(p);
+			kmem_cache_free(engine->i915->priorities, p);
 	}
 done:
 	engine->execlist_first = rb;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 626db6185a21..8529746dd7cc 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -499,7 +499,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		rb_erase(&p->node, &engine->execlist_queue);
 		INIT_LIST_HEAD(&p->requests);
 		if (p->priority != I915_PRIORITY_NORMAL)
-			kfree(p);
+			kmem_cache_free(engine->i915->priorities, p);
 	}
 done:
 	engine->execlist_first = rb;
@@ -661,7 +661,7 @@ insert_request(struct intel_engine_cs *engine,
 	if (prio == I915_PRIORITY_NORMAL) {
 		p = &engine->default_priolist;
 	} else {
-		p = kmalloc(sizeof(*p), GFP_ATOMIC);
+		p = kmem_cache_alloc(engine->i915->priorities, GFP_ATOMIC);
 		/* Convert an allocation failure to a priority bump */
 		if (unlikely(!p)) {
 			prio = I915_PRIORITY_NORMAL; /* recurses just once */
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 2c1500d0d55a..f4edd4c6cb07 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -74,6 +74,7 @@ static void mock_device_release(struct drm_device *dev)
 
 	destroy_workqueue(i915->wq);
 
+	kmem_cache_destroy(i915->priorities);
 	kmem_cache_destroy(i915->dependencies);
 	kmem_cache_destroy(i915->requests);
 	kmem_cache_destroy(i915->vmas);
@@ -186,12 +187,16 @@ struct drm_i915_private *mock_gem_device(void)
 	if (!i915->dependencies)
 		goto err_requests;
 
+	i915->priorities = KMEM_CACHE(i915_priolist, SLAB_HWCACHE_ALIGN);
+	if (!i915->priorities)
+		goto err_dependencies;
+
 	mutex_lock(&i915->drm.struct_mutex);
 	INIT_LIST_HEAD(&i915->gt.timelines);
 	err = i915_gem_timeline_init__global(i915);
 	if (err) {
 		mutex_unlock(&i915->drm.struct_mutex);
-		goto err_dependencies;
+		goto err_priorities;
 	}
 
 	mock_init_ggtt(i915);
@@ -211,6 +216,8 @@ struct drm_i915_private *mock_gem_device(void)
 err_engine:
 	for_each_engine(engine, i915, id)
 		mock_engine_free(engine);
+err_priorities:
+	kmem_cache_destroy(i915->priorities);
 err_dependencies:
 	kmem_cache_destroy(i915->dependencies);
 err_requests:
-- 
2.11.0