drm/i915/frontbuffer: Move bo refcounting intel_frontbuffer_{get,release}()

Currently xe's intel_frontbuffer implementation forgets to
hold a reference on the bo. This makes the entire thing
extremely fragile as the cleanup order now depends on bo
references held by other things
(namely intel_fb_bo_framebuffer_fini()).

Move the bo refcounting to intel_frontbuffer_{get,release}()
so that both i915 and xe do this the same way.

I first tried to fix this by having xe do the refcounting
from its intel_bo_set_frontbuffer() implementation
(which is what i915 does currently), but turns out xe's
drm_gem_object_free() can sleep and thus drm_gem_object_put()
isn't safe to call while we hold fb_tracking.lock.

Fixes: 10690b8a49bc ("drm/i915/display: Add intel_fb_bo_framebuffer_fini")
Signed-off-by: Ville Syrjälä <[e-mail]>
Link: https://patchwork.freedesktop.org/patch/msgid/[e-mail]
Reviewed-by: Jani Nikula <[e-mail]>
(cherry picked from commit eb4d490729a5fd8dc5a76d334f8d01fec7c14bbe)
Signed-off-by: Rodrigo Vivi <[e-mail]>

drivers/gpu/drm/i915/display/intel_frontbuffer.c

···
 	spin_unlock(&display->fb_tracking.lock);
 
 	i915_active_fini(&front->write);
+
+	drm_gem_object_put(obj);
 	kfree_rcu(front, rcu);
 }
 
···
 	if (!front)
 		return NULL;
 
+	drm_gem_object_get(obj);
+
 	front->obj = obj;
 	kref_init(&front->ref);
 	atomic_set(&front->bits, 0);
···
 	spin_lock(&display->fb_tracking.lock);
 	cur = intel_bo_set_frontbuffer(obj, front);
 	spin_unlock(&display->fb_tracking.lock);
-	if (cur != front)
+
+	if (cur != front) {
+		drm_gem_object_put(obj);
 		kfree(front);
+	}
+
 	return cur;
 }
 

drivers/gpu/drm/i915/gem/i915_gem_object_frontbuffer.h

···
 
 	if (!front) {
 		RCU_INIT_POINTER(obj->frontbuffer, NULL);
-		drm_gem_object_put(intel_bo_to_drm_bo(obj));
 	} else if (rcu_access_pointer(obj->frontbuffer)) {
 		cur = rcu_dereference_protected(obj->frontbuffer, true);
 		kref_get(&cur->ref);
 	} else {
-		drm_gem_object_get(intel_bo_to_drm_bo(obj));
 		rcu_assign_pointer(obj->frontbuffer, front);
 	}