[PATCH 30/30] drm/i915: Keep contexts pinned until after the next kernel context switch

Chris Wilson chris at chris-wilson.co.uk
Sat Mar 9 16:13:35 UTC 2019


We need to keep the context image pinned in memory until after the GPU
has finished writing into it. Since it continues to write as we signal
the final breadcrumb, we need to keep it pinned until the request after
it is complete. Currently we know the order in which requests execute on
each engine, and so to remove that presumption we need to identify a
request/context-switch we know must occur after our completion. Any
request queued after the signal must imply a context switch, for
simplicity we use a fresh request from the kernel context.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/gem/i915_gem_context.c   | 35 +++-----
 drivers/gpu/drm/i915/gem/i915_gem_context.h   |  1 -
 drivers/gpu/drm/i915/i915_active.c            | 72 ++++++++++++++++
 drivers/gpu/drm/i915/i915_active.h            |  4 +
 drivers/gpu/drm/i915/i915_drv.h               |  1 +
 drivers/gpu/drm/i915/i915_gem.c               | 27 ++++--
 drivers/gpu/drm/i915/i915_gem_evict.c         | 25 +++---
 drivers/gpu/drm/i915/i915_request.c           | 15 ----
 drivers/gpu/drm/i915/i915_reset.c             |  5 +-
 drivers/gpu/drm/i915/intel_context.c          | 82 +++++++++++++++++--
 drivers/gpu/drm/i915/intel_context.h          |  4 +
 drivers/gpu/drm/i915/intel_context_types.h    |  6 +-
 drivers/gpu/drm/i915/intel_engine_cs.c        | 23 +-----
 drivers/gpu/drm/i915/intel_engine_types.h     | 13 +--
 drivers/gpu/drm/i915/intel_lrc.c              | 81 ++++++------------
 drivers/gpu/drm/i915/intel_ringbuffer.c       | 59 ++-----------
 drivers/gpu/drm/i915/intel_ringbuffer.h       |  2 -
 drivers/gpu/drm/i915/selftests/mock_engine.c  |  5 --
 .../gpu/drm/i915/selftests/mock_gem_device.c  |  1 -
 19 files changed, 237 insertions(+), 224 deletions(-)

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c
index 601bd49a2b42..bde4cd830aba 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c
@@ -653,17 +653,6 @@ int i915_gem_contexts_init(struct drm_i915_private *dev_priv)
 	return 0;
 }
 
-void i915_gem_contexts_lost(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
-
-	for_each_engine(engine, dev_priv, id)
-		intel_engine_lost_context(engine);
-}
-
 void i915_gem_contexts_fini(struct drm_i915_private *i915)
 {
 	lockdep_assert_held(&i915->drm.struct_mutex);
@@ -946,6 +935,7 @@ int i915_gem_switch_to_kernel_context(struct drm_i915_private *i915,
 							 I915_FENCE_GFP);
 		}
 
+		i915_request_add_barriers(rq);
 		i915_request_add(rq);
 	}
 
@@ -1130,15 +1120,6 @@ gen8_modify_rpcs(struct intel_context *ce, struct intel_sseu sseu)
 			goto out_add;
 	}
 
-	/* Order all following requests to be after. */
-	ret = i915_timeline_set_barrier(ce->ring->timeline, rq);
-	if (ret)
-		goto out_add;
-
-	ret = gen8_emit_rpcs_config(rq, ce, sseu);
-	if (ret)
-		goto out_add;
-
 	/*
 	 * Guarantee context image and the timeline remains pinned until the
 	 * modifying request is retired by setting the ce activity tracker.
@@ -1146,9 +1127,17 @@ gen8_modify_rpcs(struct intel_context *ce, struct intel_sseu sseu)
 	 * But we only need to take one pin on the account of it. Or in other
 	 * words transfer the pinned ce object to tracked active request.
 	 */
-	if (!i915_active_request_isset(&ce->active_tracker))
-		__intel_context_pin(ce);
-	__i915_active_request_set(&ce->active_tracker, rq);
+	GEM_BUG_ON(i915_active_is_idle(&ce->active));
+	ret = i915_active_ref(&ce->active, rq->fence.context, rq);
+	if (ret)
+		goto out_add;
+
+	/* Order all following requests to be after. */
+	ret = i915_timeline_set_barrier(ce->ring->timeline, rq);
+	if (ret)
+		goto out_add;
+
+	ret = gen8_emit_rpcs_config(rq, ce, sseu);
 
 out_add:
 	i915_request_add(rq);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.h b/drivers/gpu/drm/i915/gem/i915_gem_context.h
index 107c713073cb..0e813faca451 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.h
@@ -120,7 +120,6 @@ static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx)
 
 /* i915_gem_context.c */
 int __must_check i915_gem_contexts_init(struct drm_i915_private *dev_priv);
-void i915_gem_contexts_lost(struct drm_i915_private *dev_priv);
 void i915_gem_contexts_fini(struct drm_i915_private *dev_priv);
 
 int i915_gem_context_open(struct drm_i915_private *i915,
diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c
index 863ae12707ba..e1a8758fd5d2 100644
--- a/drivers/gpu/drm/i915/i915_active.c
+++ b/drivers/gpu/drm/i915/i915_active.c
@@ -263,6 +263,78 @@ void i915_active_fini(struct i915_active *ref)
 }
 #endif
 
+int i915_active_acquire_barrier(struct i915_active *ref,
+				struct intel_engine_cs *engine)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	unsigned long tmp;
+	int err = 0;
+
+	i915_active_acquire(ref);
+	for_each_engine_masked(engine, i915, engine->mask, tmp) {
+		struct intel_context *kctx = engine->kernel_context;
+		u64 idx = kctx->ring->timeline->fence_context;
+		struct rb_node **p, *parent;
+		struct active_node *node;
+
+		parent = NULL;
+		p = &ref->tree.rb_node;
+		while (*p) {
+			parent = *p;
+
+			node = rb_entry(parent, struct active_node, node);
+			if (node->timeline == idx)
+				goto replace;
+
+			if (node->timeline < idx)
+				p = &parent->rb_right;
+			else
+				p = &parent->rb_left;
+		}
+
+		/* Think before you shrink! */
+		node = kmem_cache_alloc(global.slab_cache, GFP_KERNEL);
+		if (unlikely(!node)) {
+			err = -ENOMEM;
+			break;
+		}
+
+		i915_active_request_init(&node->base, NULL, node_retire);
+		node->ref = ref;
+		node->timeline = idx;
+
+		rb_link_node(&node->node, parent, p);
+		rb_insert_color(&node->node, &ref->tree);
+
+replace:
+		if (!node->base.request) { /* XXX recurse timeline lock */
+			RCU_INIT_POINTER(node->base.request, ERR_PTR(-EAGAIN));
+			llist_add((struct llist_node *)&node->base.link,
+				  &engine->barrier_tasks);
+			ref->count++;
+		}
+
+		__set_bit(engine->id, (unsigned long *)&i915->gt.idle_barriers);
+	}
+	i915_active_release(ref);
+
+	return err;
+}
+
+void i915_request_add_barriers(struct i915_request *rq)
+{
+	struct intel_engine_cs *engine = rq->engine;
+	struct llist_node *node, *next;
+
+	GEM_BUG_ON(!is_power_of_2(engine->mask));
+	if (!__test_and_clear_bit(engine->id,
+				  (unsigned long *)&engine->i915->gt.idle_barriers))
+		return;
+
+	llist_for_each_safe(node, next, llist_del_all(&engine->barrier_tasks))
+		list_add_tail((struct list_head *)node, &rq->active_list);
+}
+
 int i915_active_request_set(struct i915_active_request *active,
 			    struct i915_request *rq)
 {
diff --git a/drivers/gpu/drm/i915/i915_active.h b/drivers/gpu/drm/i915/i915_active.h
index 7d758719ce39..a9fd06cf1fc5 100644
--- a/drivers/gpu/drm/i915/i915_active.h
+++ b/drivers/gpu/drm/i915/i915_active.h
@@ -406,4 +406,8 @@ void i915_active_fini(struct i915_active *ref);
 static inline void i915_active_fini(struct i915_active *ref) { }
 #endif
 
+int i915_active_acquire_barrier(struct i915_active *ref,
+				struct intel_engine_cs *engine);
+void i915_request_add_barriers(struct i915_request *rq);
+
 #endif /* _I915_ACTIVE_H_ */
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index b1262e7e17b9..3e0a7b9e34ae 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1986,6 +1986,7 @@ struct drm_i915_private {
 		} timelines;
 
 		intel_engine_mask_t active_engines;
+		intel_engine_mask_t idle_barriers;
 		struct list_head active_rings;
 		u32 active_requests;
 
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 8b2c9dcba0e0..67cfb86771f3 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -80,6 +80,7 @@ static void __i915_gem_park(struct drm_i915_private *i915)
 
 	lockdep_assert_held(&i915->drm.struct_mutex);
 	GEM_BUG_ON(i915->gt.active_requests);
+	GEM_BUG_ON(i915->gt.idle_barriers);
 	GEM_BUG_ON(!list_empty(&i915->gt.active_rings));
 
 	if (!i915->gt.awake)
@@ -1020,8 +1021,8 @@ i915_gem_retire_work_handler(struct work_struct *work)
 				   round_jiffies_up_relative(HZ));
 }
 
-static bool switch_to_kernel_context_sync(struct drm_i915_private *i915,
-					  unsigned long mask)
+static bool __switch_to_kernel_context_sync(struct drm_i915_private *i915,
+					    unsigned long mask)
 {
 	bool result = true;
 
@@ -1054,10 +1055,22 @@ static bool switch_to_kernel_context_sync(struct drm_i915_private *i915,
 	return result;
 }
 
+static bool switch_to_kernel_context_idle(struct drm_i915_private *i915,
+					  unsigned long mask)
+{
+	if (!__switch_to_kernel_context_sync(i915, mask))
+		return false;
+
+	if (!__switch_to_kernel_context_sync(i915, i915->gt.idle_barriers))
+		return false;
+
+	return true;
+}
+
 static bool load_power_context(struct drm_i915_private *i915)
 {
 	/* Force loading the kernel context on all engines */
-	if (!switch_to_kernel_context_sync(i915, ALL_ENGINES))
+	if (!switch_to_kernel_context_idle(i915, ALL_ENGINES))
 		return false;
 
 	/*
@@ -1105,7 +1118,7 @@ i915_gem_idle_work_handler(struct work_struct *work)
 	    !i915->gt.active_requests) {
 		++i915->gt.active_requests; /* don't requeue idle */
 
-		switch_to_kernel_context_sync(i915, i915->gt.active_engines);
+		switch_to_kernel_context_idle(i915, i915->gt.active_engines);
 
 		if (!--i915->gt.active_requests) {
 			__i915_gem_park(i915);
@@ -1376,10 +1389,6 @@ void i915_gem_sanitize(struct drm_i915_private *i915)
 
 	intel_uncore_forcewake_put(i915, FORCEWAKE_ALL);
 	intel_runtime_pm_put(i915, wakeref);
-
-	mutex_lock(&i915->drm.struct_mutex);
-	i915_gem_contexts_lost(i915);
-	mutex_unlock(&i915->drm.struct_mutex);
 }
 
 void i915_gem_suspend(struct drm_i915_private *i915)
@@ -1404,7 +1413,7 @@ void i915_gem_suspend(struct drm_i915_private *i915)
 	 * state. Fortunately, the kernel_context is disposable and we do
 	 * not rely on its state.
 	 */
-	switch_to_kernel_context_sync(i915, i915->gt.active_engines);
+	switch_to_kernel_context_idle(i915, i915->gt.active_engines);
 
 	mutex_unlock(&i915->drm.struct_mutex);
 	i915_reset_flush(i915);
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 0947485775a3..1748501a76cb 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -40,7 +40,7 @@ I915_SELFTEST_DECLARE(static struct igt_evict_ctl {
 
 static bool ggtt_is_idle(struct drm_i915_private *i915)
 {
-	return !i915->gt.active_requests;
+	return !i915->gt.active_requests && !i915->gt.idle_barriers;
 }
 
 static int ggtt_flush(struct drm_i915_private *i915)
@@ -54,19 +54,14 @@ static int ggtt_flush(struct drm_i915_private *i915)
 	 * the hopes that we can then remove contexts and the like only
 	 * bound by their active reference.
 	 */
-	err = i915_gem_switch_to_kernel_context(i915, i915->gt.active_engines);
+	err = i915_gem_switch_to_kernel_context(i915, i915->gt.idle_barriers);
 	if (err)
 		return err;
 
-	err = i915_gem_wait_for_idle(i915,
-				     I915_WAIT_INTERRUPTIBLE |
-				     I915_WAIT_LOCKED,
-				     MAX_SCHEDULE_TIMEOUT);
-	if (err)
-		return err;
-
-	GEM_BUG_ON(!ggtt_is_idle(i915));
-	return 0;
+	return i915_gem_wait_for_idle(i915,
+				      I915_WAIT_INTERRUPTIBLE |
+				      I915_WAIT_LOCKED,
+				      MAX_SCHEDULE_TIMEOUT);
 }
 
 static bool
@@ -416,9 +411,11 @@ int i915_gem_evict_vm(struct i915_address_space *vm)
 	 * switch otherwise is ineffective.
 	 */
 	if (i915_is_ggtt(vm)) {
-		ret = ggtt_flush(vm->i915);
-		if (ret)
-			return ret;
+		do {
+			ret = ggtt_flush(vm->i915);
+			if (ret)
+				return ret;
+		} while (!ggtt_is_idle(vm->i915));
 	}
 
 	INIT_LIST_HEAD(&eviction_list);
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index b48cf1b40e81..fb24ee24c5ce 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -226,18 +226,6 @@ static void __retire_engine_request(struct intel_engine_cs *engine,
 	spin_unlock(&rq->lock);
 
 	local_irq_enable();
-
-	/*
-	 * The backing object for the context is done after switching to the
-	 * *next* context. Therefore we cannot retire the previous context until
-	 * the next context has already started running. However, since we
-	 * cannot take the required locks at i915_request_submit() we
-	 * defer the unpinning of the active context to now, retirement of
-	 * the subsequent request.
-	 */
-	if (engine->last_retired_context)
-		intel_context_unpin(engine->last_retired_context);
-	engine->last_retired_context = rq->hw_context;
 }
 
 static void __retire_engine_upto(struct intel_engine_cs *engine,
@@ -778,9 +766,6 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	if (ret)
 		goto err_unwind;
 
-	/* Keep a second pin for the dual retirement along engine and ring */
-	__intel_context_pin(ce);
-
 	rq->infix = rq->ring->emit; /* end of header; start of user payload */
 
 	/* Check that we didn't interrupt ourselves with a new request */
diff --git a/drivers/gpu/drm/i915/i915_reset.c b/drivers/gpu/drm/i915/i915_reset.c
index c56e9571cd31..f6a39cd05c5a 100644
--- a/drivers/gpu/drm/i915/i915_reset.c
+++ b/drivers/gpu/drm/i915/i915_reset.c
@@ -762,9 +762,8 @@ static void reset_restart(struct drm_i915_private *i915)
 
 	/*
 	 * Following the reset, ensure that we always reload context for
-	 * powersaving, and to correct engine->last_retired_context. Since
-	 * this requires us to submit a request, queue a worker to do that
-	 * task for us to evade any locking here.
+	 * powersaving. Since this requires us to submit a request, queue
+	 * a worker to do that task for us to evade any locking here.
 	 */
 	if (READ_ONCE(i915->gpu_error.restart))
 		return;
diff --git a/drivers/gpu/drm/i915/intel_context.c b/drivers/gpu/drm/i915/intel_context.c
index 28c702430522..a8588986406b 100644
--- a/drivers/gpu/drm/i915/intel_context.c
+++ b/drivers/gpu/drm/i915/intel_context.c
@@ -201,13 +201,16 @@ void intel_context_unpin(struct intel_context *ce)
 	mutex_unlock(&ce->pin_mutex);
 }
 
-static void intel_context_retire(struct i915_active_request *active,
-				 struct i915_request *rq)
+void intel_context_retire(struct i915_active *active)
 {
-	struct intel_context *ce =
-		container_of(active, typeof(*ce), active_tracker);
+	struct intel_context *ce = container_of(active, typeof(*ce), active);
 
-	intel_context_unpin(ce);
+	if (ce->state) {
+		ce->state->obj->pin_global--;
+		i915_vma_unpin(ce->state);
+	}
+
+	i915_gem_context_put(ce->gem_context);
 }
 
 void
@@ -227,8 +230,73 @@ intel_context_init(struct intel_context *ce,
 	/* Use the whole device by default */
 	ce->sseu = intel_device_default_sseu(ctx->i915);
 
-	i915_active_request_init(&ce->active_tracker,
-				 NULL, intel_context_retire);
+	i915_active_init(ctx->i915, &ce->active, intel_context_retire);
+}
+
+static int __context_pin_state(struct i915_vma *vma, unsigned long flags)
+{
+	int err;
+
+	/*
+	 * Clear this page out of any CPU caches for coherent swap-in/out.
+	 * We only want to do this on the first bind so that we do not stall
+	 * on an active context (which by nature is already on the GPU).
+	 */
+	if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
+		i915_gem_object_lock(vma->obj);
+		err = i915_gem_object_set_to_wc_domain(vma->obj, true);
+		i915_gem_object_unlock(vma->obj);
+		if (err)
+			return err;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, flags | PIN_GLOBAL);
+	if (err)
+		return err;
+
+	/*
+	 * And mark is as a globally pinned object to let the shrinker know
+	 * it cannot reclaim the object until we release it.
+	 */
+	vma->obj->pin_global++;
+	return 0;
+}
+
+int intel_context_active(struct intel_context *ce, unsigned long flags)
+{
+	int err;
+
+	if (!i915_active_acquire(&ce->active))
+		return 0;
+
+	if (!ce->state)
+		return 0;
+
+	err = __context_pin_state(ce->state, flags);
+	if (err) {
+		i915_active_cancel(&ce->active);
+		return err;
+	}
+
+	/* Preallocate tracking nodes */
+	if (!i915_gem_context_is_kernel(ce->gem_context)) {
+		err = i915_active_acquire_barrier(&ce->active, ce->engine);
+		if (err) {
+			i915_active_release(&ce->active);
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+void intel_context_inactive(struct intel_context *ce)
+{
+	/* Nodes preallocated in intel_context_active() */
+	if (ce->state && !i915_gem_context_is_kernel(ce->gem_context))
+		i915_active_acquire_barrier(&ce->active, ce->engine);
+
+	i915_active_release(&ce->active);
 }
 
 static void i915_global_context_shrink(void)
diff --git a/drivers/gpu/drm/i915/intel_context.h b/drivers/gpu/drm/i915/intel_context.h
index 9546d932406a..d5ead6f319a6 100644
--- a/drivers/gpu/drm/i915/intel_context.h
+++ b/drivers/gpu/drm/i915/intel_context.h
@@ -73,4 +73,8 @@ static inline void __intel_context_pin(struct intel_context *ce)
 
 void intel_context_unpin(struct intel_context *ce);
 
+int intel_context_active(struct intel_context *ce, unsigned long flags);
+void intel_context_inactive(struct intel_context *ce);
+void intel_context_retire(struct i915_active *active);
+
 #endif /* __INTEL_CONTEXT_H__ */
diff --git a/drivers/gpu/drm/i915/intel_context_types.h b/drivers/gpu/drm/i915/intel_context_types.h
index fcc030733daf..b46dc0d31ca4 100644
--- a/drivers/gpu/drm/i915/intel_context_types.h
+++ b/drivers/gpu/drm/i915/intel_context_types.h
@@ -54,10 +54,10 @@ struct intel_context {
 	struct mutex pin_mutex; /* guards pinning and associated on-gpuing */
 
 	/**
-	 * active_tracker: Active tracker for the external rq activity
-	 * on this intel_context object.
+	 * active: Active tracker for the rq activity (inc. external) on this
+	 * intel_context object.
 	 */
-	struct i915_active_request active_tracker;
+	struct i915_active active;
 
 	const struct intel_context_ops *ops;
 	struct rb_node node;
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 4eb0e879b76a..811e113f434f 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -574,6 +574,8 @@ int intel_engine_setup_common(struct intel_engine_cs *engine)
 {
 	int err;
 
+	init_llist_head(&engine->barrier_tasks);
+
 	err = init_status_page(engine);
 	if (err)
 		return err;
@@ -779,6 +781,7 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 	if (engine->preempt_context)
 		intel_context_unpin(engine->preempt_context);
 	intel_context_unpin(engine->kernel_context);
+	GEM_BUG_ON(!llist_empty(&engine->barrier_tasks));
 
 	i915_timeline_fini(&engine->timeline);
 
@@ -1210,26 +1213,6 @@ void intel_engines_unpark(struct drm_i915_private *i915)
 	}
 }
 
-/**
- * intel_engine_lost_context: called when the GPU is reset into unknown state
- * @engine: the engine
- *
- * We have either reset the GPU or otherwise about to lose state tracking of
- * the current GPU logical state (e.g. suspend). On next use, it is therefore
- * imperative that we make no presumptions about the current state and load
- * from scratch.
- */
-void intel_engine_lost_context(struct intel_engine_cs *engine)
-{
-	struct intel_context *ce;
-
-	lockdep_assert_held(&engine->i915->drm.struct_mutex);
-
-	ce = fetch_and_zero(&engine->last_retired_context);
-	if (ce)
-		intel_context_unpin(ce);
-}
-
 bool intel_engine_can_store_dword(struct intel_engine_cs *engine)
 {
 	switch (INTEL_GEN(engine->i915)) {
diff --git a/drivers/gpu/drm/i915/intel_engine_types.h b/drivers/gpu/drm/i915/intel_engine_types.h
index 525e1f1c4479..f95b276f8ad1 100644
--- a/drivers/gpu/drm/i915/intel_engine_types.h
+++ b/drivers/gpu/drm/i915/intel_engine_types.h
@@ -10,6 +10,7 @@
 #include <linux/hashtable.h>
 #include <linux/irq_work.h>
 #include <linux/list.h>
+#include <linux/llist.h>
 #include <linux/types.h>
 
 #include "i915_timeline_types.h"
@@ -267,6 +268,7 @@ struct intel_engine_cs {
 	struct intel_ring *buffer;
 
 	struct i915_timeline timeline;
+	struct llist_head barrier_tasks;
 
 	struct intel_context *kernel_context; /* pinned */
 	struct intel_context *preempt_context; /* pinned; optional */
@@ -410,17 +412,6 @@ struct intel_engine_cs {
 
 	struct intel_engine_execlists execlists;
 
-	/* Contexts are pinned whilst they are active on the GPU. The last
-	 * context executed remains active whilst the GPU is idle - the
-	 * switch away and write to the context object only occurs on the
-	 * next execution.  Contexts are only unpinned on retirement of the
-	 * following request ensuring that we can always write to the object
-	 * on the context switch even after idling. Across suspend, we switch
-	 * to the kernel context and trash it as the save may not happen
-	 * before the hardware is powered down.
-	 */
-	struct intel_context *last_retired_context;
-
 	/* status_notifier: list of callbacks for context-switch changes */
 	struct atomic_notifier_head context_status_notifier;
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 438fe374587f..94a88c097751 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1459,62 +1459,15 @@ static void execlists_context_destroy(struct intel_context *ce)
 
 static void execlists_context_unpin(struct intel_context *ce)
 {
-	struct intel_engine_cs *engine;
-
-	/*
-	 * The tasklet may still be using a pointer to our state, via an
-	 * old request. However, since we know we only unpin the context
-	 * on retirement of the following request, we know that the last
-	 * request referencing us will have had a completion CS interrupt.
-	 * If we see that it is still active, it means that the tasklet hasn't
-	 * had the chance to run yet; let it run before we teardown the
-	 * reference it may use.
-	 */
-	engine = READ_ONCE(ce->inflight);
-	if (unlikely(engine)) {
-		unsigned long flags;
-
-		spin_lock_irqsave(&engine->timeline.lock, flags);
-		process_csb(engine);
-		spin_unlock_irqrestore(&engine->timeline.lock, flags);
-
-		GEM_BUG_ON(READ_ONCE(ce->inflight));
-	}
-
 	i915_gem_context_unpin_hw_id(ce->gem_context);
 
 	intel_ring_unpin(ce->ring);
 
-	ce->state->obj->pin_global--;
 	i915_gem_object_unpin_map(ce->state->obj);
-	i915_vma_unpin(ce->state);
 
 	list_del(&ce->active_link);
-	i915_gem_context_put(ce->gem_context);
-}
-
-static int __context_pin(struct i915_vma *vma)
-{
-	unsigned int flags;
-	int err;
-
-	/*
-	 * Clear this page out of any CPU caches for coherent swap-in/out.
-	 * We only want to do this on the first bind so that we do not stall
-	 * on an active context (which by nature is already on the GPU).
-	 */
-	if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
-		i915_gem_object_lock(vma->obj);
-		err = i915_gem_object_set_to_wc_domain(vma->obj, true);
-		i915_gem_object_unlock(vma->obj);
-		if (err)
-			return err;
-	}
 
-	flags = PIN_GLOBAL | PIN_HIGH;
-	flags |= PIN_OFFSET_BIAS | i915_ggtt_pin_bias(vma);
-
-	return i915_vma_pin(vma, 0, 0, flags);
+	intel_context_inactive(ce);
 }
 
 static void
@@ -1551,7 +1504,10 @@ __execlists_context_pin(struct intel_context *ce,
 		goto err;
 	GEM_BUG_ON(!ce->state);
 
-	ret = __context_pin(ce->state);
+	ret = intel_context_active(ce,
+				   engine->i915->ggtt.pin_bias |
+				   PIN_OFFSET_BIAS |
+				   PIN_HIGH);
 	if (ret)
 		goto err;
 
@@ -1560,7 +1516,7 @@ __execlists_context_pin(struct intel_context *ce,
 					I915_MAP_OVERRIDE);
 	if (IS_ERR(vaddr)) {
 		ret = PTR_ERR(vaddr);
-		goto unpin_vma;
+		goto unpin_active;
 	}
 
 	ret = intel_ring_pin(ce->ring);
@@ -1575,15 +1531,14 @@ __execlists_context_pin(struct intel_context *ce,
 	ce->lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
 	__execlists_update_reg_state(ce, engine);
 
-	ce->state->obj->pin_global++;
 	return 0;
 
 unpin_ring:
 	intel_ring_unpin(ce->ring);
 unpin_map:
 	i915_gem_object_unpin_map(ce->state->obj);
-unpin_vma:
-	__i915_vma_unpin(ce->state);
+unpin_active:
+	intel_context_inactive(ce);
 err:
 	return ret;
 }
@@ -3151,13 +3106,20 @@ static void virtual_engine_free(struct kref *kref)
 	call_rcu(&ve->rcu, __virtual_engine_free);
 }
 
+static void virtual_context_retire(struct i915_active *ref)
+{
+	struct virtual_engine *ve =
+		container_of(ref, typeof(*ve), context.active);
+
+	intel_context_retire(ref);
+	kref_put(&ve->kref, virtual_engine_free);
+}
+
 static void virtual_context_unpin(struct intel_context *ce)
 {
 	struct virtual_engine *ve = container_of(ce, typeof(*ve), context);
 
 	execlists_context_unpin(ce);
-
-	kref_put(&ve->kref, virtual_engine_free);
 }
 
 static void virtual_engine_initial_hint(struct virtual_engine *ve)
@@ -3191,14 +3153,18 @@ static int virtual_context_pin(struct intel_context *ce)
 	struct virtual_engine *ve = container_of(ce, typeof(*ve), context);
 	int err;
 
+	if (i915_active_is_idle(&ce->active))
+		kref_get(&ve->kref);
+
 	/* Note: we must use a real engine class for setting up reg state */
 	err = __execlists_context_pin(ce, ve->siblings[0]);
-	if (err)
+	if (err) {
+		kref_put(&ve->kref, virtual_engine_free);
 		return err;
+	}
 
 	virtual_engine_initial_hint(ve);
 
-	kref_get(&ve->kref);
 	return 0;
 }
 
@@ -3372,6 +3338,7 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 		     (unsigned long)ve);
 
 	intel_context_init(&ve->context, ctx, &ve->base);
+	ve->context.active.retire = virtual_context_retire;
 
 	for (n = 0; n < count; n++) {
 		struct intel_engine_cs *sibling = siblings[n];
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 20aff03ccb94..847c600f5d19 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1389,60 +1389,13 @@ static void __context_unpin_ppgtt(struct i915_gem_context *ctx)
 		gen6_ppgtt_unpin(ppgtt);
 }
 
-static int __context_pin(struct intel_context *ce)
-{
-	struct i915_vma *vma;
-	int err;
-
-	vma = ce->state;
-	if (!vma)
-		return 0;
-
-	/*
-	 * Clear this page out of any CPU caches for coherent swap-in/out.
-	 * We only want to do this on the first bind so that we do not stall
-	 * on an active context (which by nature is already on the GPU).
-	 */
-	if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
-		i915_gem_object_lock(vma->obj);
-		err = i915_gem_object_set_to_gtt_domain(vma->obj, true);
-		i915_gem_object_unlock(vma->obj);
-		if (err)
-			return err;
-	}
-
-	err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
-	if (err)
-		return err;
-
-	/*
-	 * And mark is as a globally pinned object to let the shrinker know
-	 * it cannot reclaim the object until we release it.
-	 */
-	vma->obj->pin_global++;
-
-	return 0;
-}
-
-static void __context_unpin(struct intel_context *ce)
-{
-	struct i915_vma *vma;
-
-	vma = ce->state;
-	if (!vma)
-		return;
-
-	vma->obj->pin_global--;
-	i915_vma_unpin(vma);
-}
-
 static void ring_context_unpin(struct intel_context *ce)
 {
 	__context_unpin_ppgtt(ce->gem_context);
-	__context_unpin(ce);
 
 	list_del(&ce->active_link);
-	i915_gem_context_put(ce->gem_context);
+
+	intel_context_inactive(ce);
 }
 
 static struct i915_vma *
@@ -1531,18 +1484,18 @@ static int ring_context_pin(struct intel_context *ce)
 		ce->state = vma;
 	}
 
-	err = __context_pin(ce);
+	err = intel_context_active(ce, PIN_HIGH);
 	if (err)
 		return err;
 
 	err = __context_pin_ppgtt(ce->gem_context);
 	if (err)
-		goto err_unpin;
+		goto err_active;
 
 	return 0;
 
-err_unpin:
-	__context_unpin(ce);
+err_active:
+	intel_context_inactive(ce);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index e612bdca9fd9..1a98a289ef45 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -435,8 +435,6 @@ void intel_engines_sanitize(struct drm_i915_private *i915, bool force);
 bool intel_engine_is_idle(struct intel_engine_cs *engine);
 bool intel_engines_are_idle(struct drm_i915_private *dev_priv);
 
-void intel_engine_lost_context(struct intel_engine_cs *engine);
-
 void intel_engines_park(struct drm_i915_private *i915);
 void intel_engines_unpark(struct drm_i915_private *i915);
 
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index bae88d5f4fbf..afda321b4941 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -271,14 +271,9 @@ void mock_engine_free(struct intel_engine_cs *engine)
 {
 	struct mock_engine *mock =
 		container_of(engine, typeof(*mock), base);
-	struct intel_context *ce;
 
 	GEM_BUG_ON(timer_pending(&mock->hw_delay));
 
-	ce = fetch_and_zero(&engine->last_retired_context);
-	if (ce)
-		intel_context_unpin(ce);
-
 	intel_context_unpin(engine->kernel_context);
 
 	intel_engine_fini_breadcrumbs(engine);
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 37c9a5706f69..b851511fcb00 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -56,7 +56,6 @@ static void mock_device_release(struct drm_device *dev)
 
 	mutex_lock(&i915->drm.struct_mutex);
 	mock_device_flush(i915);
-	i915_gem_contexts_lost(i915);
 	mutex_unlock(&i915->drm.struct_mutex);
 
 	drain_delayed_work(&i915->gt.retire_work);
-- 
2.20.1



More information about the Intel-gfx-trybot mailing list