1 files changed, 209 insertions, 110 deletions

diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c
index 52db2bde44a3..92a3f25c4006 100644
--- a/drivers/gpu/drm/i915/gt/intel_context.c
+++ b/drivers/gpu/drm/i915/gt/intel_context.c
@@ -93,57 +93,210 @@ static void intel_context_active_release(struct intel_context *ce)
 	i915_active_release(&ce->active);
 }
 
-int __intel_context_do_pin(struct intel_context *ce)
+static int __context_pin_state(struct i915_vma *vma, struct i915_gem_ww_ctx *ww)
+{
+	unsigned int bias = i915_ggtt_pin_bias(vma) | PIN_OFFSET_BIAS;
+	int err;
+
+	err = i915_ggtt_pin(vma, ww, 0, bias | PIN_HIGH);
+	if (err)
+		return err;
+
+	err = i915_active_acquire(&vma->active);
+	if (err)
+		goto err_unpin;
+
+	/*
+	 * And mark it as a globally pinned object to let the shrinker know
+	 * it cannot reclaim the object until we release it.
+	 */
+	i915_vma_make_unshrinkable(vma);
+	vma->obj->mm.dirty = true;
+
+	return 0;
+
+err_unpin:
+	i915_vma_unpin(vma);
+	return err;
+}
+
+static void __context_unpin_state(struct i915_vma *vma)
+{
+	i915_vma_make_shrinkable(vma);
+	i915_active_release(&vma->active);
+	__i915_vma_unpin(vma);
+}
+
+static int __ring_active(struct intel_ring *ring,
+			 struct i915_gem_ww_ctx *ww)
+{
+	int err;
+
+	err = intel_ring_pin(ring, ww);
+	if (err)
+		return err;
+
+	err = i915_active_acquire(&ring->vma->active);
+	if (err)
+		goto err_pin;
+
+	return 0;
+
+err_pin:
+	intel_ring_unpin(ring);
+	return err;
+}
+
+static void __ring_retire(struct intel_ring *ring)
+{
+	i915_active_release(&ring->vma->active);
+	intel_ring_unpin(ring);
+}
+
+static int intel_context_pre_pin(struct intel_context *ce,
+				 struct i915_gem_ww_ctx *ww)
 {
 	int err;
 
+	CE_TRACE(ce, "active\n");
+
+	err = __ring_active(ce->ring, ww);
+	if (err)
+		return err;
+
+	err = intel_timeline_pin(ce->timeline, ww);
+	if (err)
+		goto err_ring;
+
+	if (!ce->state)
+		return 0;
+
+	err = __context_pin_state(ce->state, ww);
+	if (err)
+		goto err_timeline;
+
+
+	return 0;
+
+err_timeline:
+	intel_timeline_unpin(ce->timeline);
+err_ring:
+	__ring_retire(ce->ring);
+	return err;
+}
+
+static void intel_context_post_unpin(struct intel_context *ce)
+{
+	if (ce->state)
+		__context_unpin_state(ce->state);
+
+	intel_timeline_unpin(ce->timeline);
+	__ring_retire(ce->ring);
+}
+
+int __intel_context_do_pin_ww(struct intel_context *ce,
+			      struct i915_gem_ww_ctx *ww)
+{
+	bool handoff = false;
+	void *vaddr;
+	int err = 0;
+
 	if (unlikely(!test_bit(CONTEXT_ALLOC_BIT, &ce->flags))) {
 		err = intel_context_alloc_state(ce);
 		if (err)
 			return err;
 	}
 
-	err = i915_active_acquire(&ce->active);
+	/*
+	 * We always pin the context/ring/timeline here, to ensure a pin
+	 * refcount for __intel_context_active(), which prevent a lock
+	 * inversion of ce->pin_mutex vs dma_resv_lock().
+	 */
+
+	err = i915_gem_object_lock(ce->timeline->hwsp_ggtt->obj, ww);
+	if (!err && ce->ring->vma->obj)
+		err = i915_gem_object_lock(ce->ring->vma->obj, ww);
+	if (!err && ce->state)
+		err = i915_gem_object_lock(ce->state->obj, ww);
+	if (!err)
+		err = intel_context_pre_pin(ce, ww);
 	if (err)
 		return err;
 
-	if (mutex_lock_interruptible(&ce->pin_mutex)) {
-		err = -EINTR;
-		goto out_release;
-	}
+	err = i915_active_acquire(&ce->active);
+	if (err)
+		goto err_ctx_unpin;
+
+	err = ce->ops->pre_pin(ce, ww, &vaddr);
+	if (err)
+		goto err_release;
+
+	err = mutex_lock_interruptible(&ce->pin_mutex);
+	if (err)
+		goto err_post_unpin;
 
 	if (unlikely(intel_context_is_closed(ce))) {
 		err = -ENOENT;
-		goto out_unlock;
+		goto err_unlock;
 	}
 
 	if (likely(!atomic_add_unless(&ce->pin_count, 1, 0))) {
 		err = intel_context_active_acquire(ce);
 		if (unlikely(err))
-			goto out_unlock;
+			goto err_unlock;
 
-		err = ce->ops->pin(ce);
-		if (unlikely(err))
-			goto err_active;
+		err = ce->ops->pin(ce, vaddr);
+		if (err) {
+			intel_context_active_release(ce);
+			goto err_unlock;
+		}
 
 		CE_TRACE(ce, "pin ring:{start:%08x, head:%04x, tail:%04x}\n",
 			 i915_ggtt_offset(ce->ring->vma),
 			 ce->ring->head, ce->ring->tail);
 
+		handoff = true;
 		smp_mb__before_atomic(); /* flush pin before it is visible */
 		atomic_inc(&ce->pin_count);
 	}
 
 	GEM_BUG_ON(!intel_context_is_pinned(ce)); /* no overflow! */
-	GEM_BUG_ON(i915_active_is_idle(&ce->active));
-	goto out_unlock;
 
-err_active:
-	intel_context_active_release(ce);
-out_unlock:
+err_unlock:
 	mutex_unlock(&ce->pin_mutex);
-out_release:
+err_post_unpin:
+	if (!handoff)
+		ce->ops->post_unpin(ce);
+err_release:
 	i915_active_release(&ce->active);
+err_ctx_unpin:
+	intel_context_post_unpin(ce);
+
+	/*
+	 * Unlock the hwsp_ggtt object since it's shared.
+	 * In principle we can unlock all the global state locked above
+	 * since it's pinned and doesn't need fencing, and will
+	 * thus remain resident until it is explicitly unpinned.
+	 */
+	i915_gem_ww_unlock_single(ce->timeline->hwsp_ggtt->obj);
+
+	return err;
+}
+
+int __intel_context_do_pin(struct intel_context *ce)
+{
+	struct i915_gem_ww_ctx ww;
+	int err;
+
+	i915_gem_ww_ctx_init(&ww, true);
+retry:
+	err = __intel_context_do_pin_ww(ce, &ww);
+	if (err == -EDEADLK) {
+		err = i915_gem_ww_ctx_backoff(&ww);
+		if (!err)
+			goto retry;
+	}
+	i915_gem_ww_ctx_fini(&ww);
 	return err;
 }
 
@@ -154,6 +307,7 @@ void intel_context_unpin(struct intel_context *ce)
 
 	CE_TRACE(ce, "unpin\n");
 	ce->ops->unpin(ce);
+	ce->ops->post_unpin(ce);
 
 	/*
 	 * Once released, we may asynchronously drop the active reference.
@@ -166,65 +320,6 @@ void intel_context_unpin(struct intel_context *ce)
 	intel_context_put(ce);
 }
 
-static int __context_pin_state(struct i915_vma *vma)
-{
-	unsigned int bias = i915_ggtt_pin_bias(vma) | PIN_OFFSET_BIAS;
-	int err;
-
-	err = i915_ggtt_pin(vma, 0, bias | PIN_HIGH);
-	if (err)
-		return err;
-
-	err = i915_active_acquire(&vma->active);
-	if (err)
-		goto err_unpin;
-
-	/*
-	 * And mark it as a globally pinned object to let the shrinker know
-	 * it cannot reclaim the object until we release it.
-	 */
-	i915_vma_make_unshrinkable(vma);
-	vma->obj->mm.dirty = true;
-
-	return 0;
-
-err_unpin:
-	i915_vma_unpin(vma);
-	return err;
-}
-
-static void __context_unpin_state(struct i915_vma *vma)
-{
-	i915_vma_make_shrinkable(vma);
-	i915_active_release(&vma->active);
-	__i915_vma_unpin(vma);
-}
-
-static int __ring_active(struct intel_ring *ring)
-{
-	int err;
-
-	err = intel_ring_pin(ring);
-	if (err)
-		return err;
-
-	err = i915_active_acquire(&ring->vma->active);
-	if (err)
-		goto err_pin;
-
-	return 0;
-
-err_pin:
-	intel_ring_unpin(ring);
-	return err;
-}
-
-static void __ring_retire(struct intel_ring *ring)
-{
-	i915_active_release(&ring->vma->active);
-	intel_ring_unpin(ring);
-}
-
 __i915_active_call
 static void __intel_context_retire(struct i915_active *active)
 {
@@ -235,48 +330,29 @@ static void __intel_context_retire(struct i915_active *active)
 		 intel_context_get_avg_runtime_ns(ce));
 
 	set_bit(CONTEXT_VALID_BIT, &ce->flags);
-	if (ce->state)
-		__context_unpin_state(ce->state);
-
-	intel_timeline_unpin(ce->timeline);
-	__ring_retire(ce->ring);
-
+	intel_context_post_unpin(ce);
 	intel_context_put(ce);
 }
 
 static int __intel_context_active(struct i915_active *active)
 {
 	struct intel_context *ce = container_of(active, typeof(*ce), active);
-	int err;
-
-	CE_TRACE(ce, "active\n");
 
 	intel_context_get(ce);
 
-	err = __ring_active(ce->ring);
-	if (err)
-		goto err_put;
+	/* everything should already be activated by intel_context_pre_pin() */
+	GEM_WARN_ON(!i915_active_acquire_if_busy(&ce->ring->vma->active));
+	__intel_ring_pin(ce->ring);
 
-	err = intel_timeline_pin(ce->timeline);
-	if (err)
-		goto err_ring;
+	__intel_timeline_pin(ce->timeline);
 
-	if (!ce->state)
-		return 0;
-
-	err = __context_pin_state(ce->state);
-	if (err)
-		goto err_timeline;
+	if (ce->state) {
+		GEM_WARN_ON(!i915_active_acquire_if_busy(&ce->state->active));
+		__i915_vma_pin(ce->state);
+		i915_vma_make_unshrinkable(ce->state);
+	}
 
 	return 0;
-
-err_timeline:
-	intel_timeline_unpin(ce->timeline);
-err_ring:
-	__ring_retire(ce->ring);
-err_put:
-	intel_context_put(ce);
-	return err;
 }
 
 void
@@ -382,15 +458,38 @@ int intel_context_prepare_remote_request(struct intel_context *ce,
 
 struct i915_request *intel_context_create_request(struct intel_context *ce)
 {
+	struct i915_gem_ww_ctx ww;
 	struct i915_request *rq;
 	int err;
 
-	err = intel_context_pin(ce);
-	if (unlikely(err))
-		return ERR_PTR(err);
+	i915_gem_ww_ctx_init(&ww, true);
+retry:
+	err = intel_context_pin_ww(ce, &ww);
+	if (!err) {
+		rq = i915_request_create(ce);
+		intel_context_unpin(ce);
+	} else if (err == -EDEADLK) {
+		err = i915_gem_ww_ctx_backoff(&ww);
+		if (!err)
+			goto retry;
+		rq = ERR_PTR(err);
+	} else {
+		rq = ERR_PTR(err);
+	}
+
+	i915_gem_ww_ctx_fini(&ww);
 
-	rq = i915_request_create(ce);
-	intel_context_unpin(ce);
+	if (IS_ERR(rq))
+		return rq;
+
+	/*
+	 * timeline->mutex should be the inner lock, but is used as outer lock.
+	 * Hack around this to shut up lockdep in selftests..
+	 */
+	lockdep_unpin_lock(&ce->timeline->mutex, rq->cookie);
+	mutex_release(&ce->timeline->mutex.dep_map, _RET_IP_);
+	mutex_acquire(&ce->timeline->mutex.dep_map, SINGLE_DEPTH_NESTING, 0, _RET_IP_);
+	rq->cookie = lockdep_pin_lock(&ce->timeline->mutex);
 
 	return rq;
 }