[PATCH v7 4/5] drm: serialize drm_file.master with a master lock

[Desmond Cheong Zhi Xi]

Currently, drm_file.master pointers should be protected by
drm_device.master_mutex when being dereferenced. This is because
drm_file.master is not invariant for the lifetime of drm_file. If
drm_file is not the creator of master, then drm_file.is_master is
false, and a call to drm_setmaster_ioctl will invoke
drm_new_set_master, which then allocates a new master for drm_file and
puts the old master.

Thus, without holding drm_device.master_mutex, the old value of
drm_file.master could be freed while it is being used by another
concurrent process.

However, it is not always possible to lock drm_device.master_mutex to
dereference drm_file.master. Through the fbdev emulation code, this
might occur in a deep nest of other locks. But drm_device.master_mutex
is also the outermost lock in the nesting hierarchy, so this leads to
potential deadlocks.

To address this, we introduce a new mutex at the bottom of the lock
hierarchy that only serializes drm_file.master. With this change, the
value of drm_file.master changes only when both
drm_device.master_mutex and drm_file.master_lock are held. Hence, any
process holding either of those locks can ensure that the value of
drm_file.master will not change concurrently.

Since no lock depends on the new drm_file.master_lock, when
drm_file.master is dereferenced, but drm_device.master_mutex cannot be
held, we can safely protect the master pointer with
drm_file.master_lock.

Reported-by: Daniel Vetter <daniel.vetter@xxxxxxxx>
Signed-off-by: Desmond Cheong Zhi Xi <desmondcheongzx@xxxxxxxxx>
---

Since our lock inversions were a result of dev->master_mutex being
used to serialize many other things, perhaps a finer grained lock will
solve the lockdep issues.

 drivers/gpu/drm/drm_auth.c | 10 ++++++++--
 drivers/gpu/drm/drm_file.c |  1 +
 include/drm/drm_file.h     | 12 +++++++++---
 3 files changed, 18 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/drm_auth.c b/drivers/gpu/drm/drm_auth.c
index ab1863c5a5a0..fe5b6adc6133 100644
--- a/drivers/gpu/drm/drm_auth.c
+++ b/drivers/gpu/drm/drm_auth.c
@@ -169,11 +169,14 @@ static int drm_new_set_master(struct drm_device *dev, struct drm_file *fpriv)
 
 	WARN_ON(fpriv->is_master);
 	old_master = fpriv->master;
+	mutex_lock(&fpriv->master_lock);
 	fpriv->master = drm_master_create(dev);
 	if (!fpriv->master) {
 		fpriv->master = old_master;
+		mutex_unlock(&fpriv->master_lock);
 		return -ENOMEM;
 	}
+	mutex_unlock(&fpriv->master_lock);
 
 	fpriv->is_master = 1;
 	fpriv->authenticated = 1;
@@ -332,10 +335,13 @@ int drm_master_open(struct drm_file *file_priv)
 	 * any master object for render clients
 	 */
 	mutex_lock(&dev->master_mutex);
-	if (!dev->master)
+	if (!dev->master) {
 		ret = drm_new_set_master(dev, file_priv);
-	else
+	} else {
+		mutex_lock(&file_priv->master_lock);
 		file_priv->master = drm_master_get(dev->master);
+		mutex_unlock(&file_priv->master_lock);
+	}
 	mutex_unlock(&dev->master_mutex);
 
 	return ret;
diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c
index d4f0bac6f8f8..8ccadfa1c752 100644
--- a/drivers/gpu/drm/drm_file.c
+++ b/drivers/gpu/drm/drm_file.c
@@ -176,6 +176,7 @@ struct drm_file *drm_file_alloc(struct drm_minor *minor)
 	init_waitqueue_head(&file->event_wait);
 	file->event_space = 4096; /* set aside 4k for event buffer */
 
+	mutex_init(&file->master_lock);
 	mutex_init(&file->event_read_lock);
 
 	if (drm_core_check_feature(dev, DRIVER_GEM))
diff --git a/include/drm/drm_file.h b/include/drm/drm_file.h
index b81b3bfb08c8..88539f93fc8e 100644
--- a/include/drm/drm_file.h
+++ b/include/drm/drm_file.h
@@ -226,15 +226,21 @@ struct drm_file {
 	/**
 	 * @master:
 	 *
-	 * Master this node is currently associated with. Only relevant if
-	 * drm_is_primary_client() returns true. Note that this only
-	 * matches &drm_device.master if the master is the currently active one.
+	 * Master this node is currently associated with. Protected by struct
+	 * &drm_device.master_mutex, and serialized by @master_lock.
+	 *
+	 * Only relevant if drm_is_primary_client() returns true. Note that
+	 * this only matches &drm_device.master if the master is the currently
+	 * active one.
 	 *
 	 * See also @authentication and @is_master and the :ref:`section on
 	 * primary nodes and authentication <drm_primary_node>`.
 	 */
 	struct drm_master *master;
 
+	/** @master_lock: Serializes @master. */
+	struct mutex master_lock;
+
 	/** @pid: Process that opened this file. */
 	struct pid *pid;
 
-- 
2.25.1