mirror of https://lore.kernel.org/linux-amlogic/
 help / color / mirror / Atom feed
* [PATCH] drm/meson: fix device_node leak in meson_encoder_dsi_probe()
@ 2026-09-06 19:43 Miles Krause via B4 Relay
  2026-09-06 19:55 ` sashiko-bot
  2026-09-10  7:23 ` Neil Armstrong
  0 siblings, 2 replies; 3+ messages in thread
From: Miles Krause via B4 Relay @ 2026-09-06 19:43 UTC (permalink / raw)
  To: Neil Armstrong, Maarten Lankhorst, Maxime Ripard,
	Thomas Zimmermann, David Airlie, Simona Vetter, Kevin Hilman,
	Jerome Brunet, Martin Blumenstingl, Nicolas Belin, Jagan Teki
  Cc: dri-devel, linux-amlogic, linux-arm-kernel, linux-kernel, Miles Krause

From: Miles Krause <mileskrause5200@gmail.com>

meson_encoder_dsi_probe() looks up the DSI transceiver's device node
with of_graph_get_remote_node() and hands it to
of_drm_find_and_get_bridge(), but never drops the reference that the
lookup returned.

of_graph_get_remote_node() returns the node with its refcount
incremented, and of_drm_find_and_get_bridge() takes its own reference
on the bridge it finds, so 'remote' is leaked on every path once the
lookup succeeds: on the -EPROBE_DEFER return taken when no bridge is
found, and on a fully successful probe alike.

The two sibling encoders in this driver already handle this correctly.
meson_encoder_cvbs_probe() calls of_node_put(remote) immediately after
of_drm_find_and_get_bridge(), and meson_encoder_hdmi_probe() releases
the node both on its error path and after of_find_device_by_node().

Release the reference in the same place meson_encoder_cvbs_probe()
does, which covers both the error and the success path.

Fixes: 42dcf15f901c ("drm/meson: add DSI encoder")
Signed-off-by: Miles Krause <mileskrause5200@gmail.com>
---
 drivers/gpu/drm/meson/meson_encoder_dsi.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/meson/meson_encoder_dsi.c b/drivers/gpu/drm/meson/meson_encoder_dsi.c
index 3e422b612f74..3fe604496294 100644
--- a/drivers/gpu/drm/meson/meson_encoder_dsi.c
+++ b/drivers/gpu/drm/meson/meson_encoder_dsi.c
@@ -120,6 +120,7 @@ int meson_encoder_dsi_probe(struct meson_drm *priv)
 	}
 
 	meson_encoder_dsi->bridge.next_bridge = of_drm_find_and_get_bridge(remote);
+	of_node_put(remote);
 	if (!meson_encoder_dsi->bridge.next_bridge)
 		return dev_err_probe(priv->dev, -EPROBE_DEFER,
 				     "Failed to find DSI transceiver bridge\n");

---
base-commit: 88405f0ad1d5c680afe3ea0ce9345fa9e1deaac8
change-id: 20260906-meson-dsi-of-node-leak-744899da5712

Best regards,
-- 
Miles Krause <mileskrause5200@gmail.com>



_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2026-09-10  7:23 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-09-06 19:43 [PATCH] drm/meson: fix device_node leak in meson_encoder_dsi_probe() Miles Krause via B4 Relay
2026-09-06 19:55 ` sashiko-bot
2026-09-10  7:23 ` Neil Armstrong

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®