The function samsung_dsim_parse_dt() calls of_graph_get_endpoint_by_regs()
to get the endpoint device node, but fails to call of_node_put() to release
the reference when the function returns. This results in a device node
reference leak.

Fix this by adding the missing of_node_put() call before returning from
the function.

Found via static analysis and code review.

Fixes: 77169a11d4e9 ("drm/bridge: samsung-dsim: add driver support for exynos7870 DSIM bridge")
Cc: stable@vger.kernel.org
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
---
 drivers/gpu/drm/bridge/samsung-dsim.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/bridge/samsung-dsim.c b/drivers/gpu/drm/bridge/samsung-dsim.c
index eabc4c32f6ab..1a5acd5077ad 100644
--- a/drivers/gpu/drm/bridge/samsung-dsim.c
+++ b/drivers/gpu/drm/bridge/samsung-dsim.c
@@ -2086,6 +2086,7 @@ static int samsung_dsim_parse_dt(struct samsung_dsim *dsi)
 		if (lane_polarities[1])
 			dsi->swap_dn_dp_data = true;
 	}
+	of_node_put(endpoint);
 
 	return 0;
 }
-- 
2.39.5 (Apple Git-154)

On 2025-10-29 07:41, Miaoqian Lin wrote:
> The function samsung_dsim_parse_dt() calls of_graph_get_endpoint_by_regs()
> to get the endpoint device node, but fails to call of_node_put() to release
> the reference when the function returns. This results in a device node
> reference leak.
> 
> Fix this by adding the missing of_node_put() call before returning from
> the function.
> 
> Found via static analysis and code review.
> 
> Fixes: 77169a11d4e9 ("drm/bridge: samsung-dsim: add driver support for exynos7870 DSIM bridge")

Is the Fixes: tag correct? This is what I get for relevant code:

74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2074)     endpoint = of_graph_get_endpoint_by_regs(node, 1, -1);
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2075)     nr_lanes = of_property_count_u32_elems(endpoint, "data-lanes");
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2076)     if (nr_lanes > 0 && nr_lanes <= 4) {
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2077)             /* Polarity 0 is clock lane, 1..4 are data lanes. */
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2078)             of_property_read_u32_array(endpoint, "lane-polarities",
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2079)                                        lane_polarities, nr_lanes + 1);
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2080)             for (i = 1; i <= nr_lanes; i++) {
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2081)                     if (lane_polarities[1] != lane_polarities[i])
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2082)                             DRM_DEV_ERROR(dsi->dev, "Data lanes polarities do not match");
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2083)             }
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2084)             if (lane_polarities[0])
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2085)                     dsi->swap_dn_dp_clk = true;
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2086)             if (lane_polarities[1])
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2087)                     dsi->swap_dn_dp_data = true;
74629c49e66cc (Marek Vasut          2023-05-14 08:46:25 -0300 2088)     }

This should be a fix for 74629c49e66c instead.