[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20251003-drm-bridge-alloc-encoder-chain-mutex-v2-2-78bf61580a06@bootlin.com>
Date: Fri, 03 Oct 2025 12:39:24 +0200
From: Luca Ceresoli <luca.ceresoli@...tlin.com>
To: Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
Maxime Ripard <mripard@...nel.org>, Thomas Zimmermann <tzimmermann@...e.de>,
David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
Andrzej Hajda <andrzej.hajda@...el.com>,
Neil Armstrong <neil.armstrong@...aro.org>, Robert Foss <rfoss@...nel.org>,
Laurent Pinchart <Laurent.pinchart@...asonboard.com>,
Jonas Karlman <jonas@...boo.se>, Jernej Skrabec <jernej.skrabec@...il.com>
Cc: Hui Pu <Hui.Pu@...ealthcare.com>,
Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
dri-devel@...ts.freedesktop.org, linux-kernel@...r.kernel.org,
Luca Ceresoli <luca.ceresoli@...tlin.com>
Subject: [PATCH v2 2/7] drm/encoder: drm_encoder_cleanup: take chain mutex
while tearing down
drm_encoder_cleanup() modifies the encoder chain by removing bridges via
drm_bridge_detach(). Protect this whole operation by taking the mutex, so
that:
* any users iterating over the chain will not access it during the change
* any other code willing to modify the list (e.g. drm_bridge_attach())
will wait until drm_encoder_cleanup() is done
Note that the _safe macro in use here is providing a different kind of
protection than the mutex:
1. list_for_each_entry_safe() allows removing the current entry from the
list it is iterating on, synchronously; the non-safe version would be
unable to find the next entry when the current entry has been removed
2. the mutex being added allows to ensure that the list is not used
concurrently by other code while it is being modified; this prevents
such other concurrent code to derail because it is iterating over an
element while it is removed
The _safe macro, which works by taking the "next" pointer in addition to
the "current" one, is insufficient to provide the protection at item
2. This is visible e.g. when the "next" element is removed by other
concurrent code. This is what would happen without the added mutex:
1. start loop: list_for_each_entry_safe(pos, n, ...) sets;
pos = list_first_entry() = (bridge 1)
n = list_next_entry(pos) = (bridge 2)
2. enter the loop 1st time, do something with *pos (bridge 1)
3. in the meanwhile bridge 2 is hot-unplugged
-> another thread removes bridge 2
-> drm_bridge_detach()
-> list_del() sets (bridge 2)->next = LIST_POISON1
4. loop iteration 1 finishes, list_for_each_entry_safe() sets:
pos = n (previously set to bridge 2)
n = (bridge 2)->next = LIST_POISON1
5. enter the loop 2nd time, do something with *pos (bridge 2)
6. loop iteration 2 finishes, list_for_each_entry_safe() sets:
pos = n = LIST_POISON1 ==> bug!
Signed-off-by: Luca Ceresoli <luca.ceresoli@...tlin.com>
---
Changes in v2:
- Expanded commit messge with rationale, as discussed
---
drivers/gpu/drm/drm_encoder.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/drm_encoder.c b/drivers/gpu/drm/drm_encoder.c
index 3261f142baea30c516499d23dbf8d0acf5952cd6..3a04bedf9b759acd6826864b7f2cc9b861a8f170 100644
--- a/drivers/gpu/drm/drm_encoder.c
+++ b/drivers/gpu/drm/drm_encoder.c
@@ -195,9 +195,11 @@ void drm_encoder_cleanup(struct drm_encoder *encoder)
* the indices on the drm_encoder after us in the encoder_list.
*/
+ mutex_lock(&encoder->bridge_chain_mutex);
list_for_each_entry_safe(bridge, next, &encoder->bridge_chain,
chain_node)
drm_bridge_detach(bridge);
+ mutex_unlock(&encoder->bridge_chain_mutex);
drm_mode_object_unregister(dev, &encoder->base);
kfree(encoder->name);
--
2.51.0
Powered by blists - more mailing lists