[PATCH] drm/sti: hda: Fix bridge leak on component_add() failure

Myeonghun Pak posted 1 patch 1 month, 3 weeks ago
drivers/gpu/drm/sti/sti_hda.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
[PATCH] drm/sti: hda: Fix bridge leak on component_add() failure
Posted by Myeonghun Pak 1 month, 3 weeks ago
sti_hda_probe() registers its bridge before adding the component. If
component_add() fails, probe returns the error directly and leaves the
bridge registered.

The remove callback unregisters the bridge, but remove is only called
after a successful probe. Remove the bridge locally when component_add()
fails so the probe error path matches the successful-probe remove path.

Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
---
 drivers/gpu/drm/sti/sti_hda.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/sti/sti_hda.c b/drivers/gpu/drm/sti/sti_hda.c
index b739782788..9223135c93 100644
--- a/drivers/gpu/drm/sti/sti_hda.c
+++ b/drivers/gpu/drm/sti/sti_hda.c
@@ -741,6 +741,7 @@ static int sti_hda_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct sti_hda *hda;
 	struct resource *res;
+	int ret;
 
 	DRM_INFO("%s\n", __func__);
 
@@ -783,7 +784,11 @@ static int sti_hda_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, hda);
 
-	return component_add(&pdev->dev, &sti_hda_ops);
+	ret = component_add(&pdev->dev, &sti_hda_ops);
+	if (ret)
+		drm_bridge_remove(&hda->bridge);
+
+	return ret;
 }
 
 static void sti_hda_remove(struct platform_device *pdev)
--
Re: [PATCH] drm/sti: hda: Fix bridge leak on component_add() failure
Posted by Luca Ceresoli 1 month, 3 weeks ago
Hello Myeonghun,

On Fri Apr 24, 2026 at 2:11 PM CEST, Myeonghun Pak wrote:
> sti_hda_probe() registers its bridge before adding the component. If
> component_add() fails, probe returns the error directly and leaves the
> bridge registered.
>
> The remove callback unregisters the bridge, but remove is only called
> after a successful probe. Remove the bridge locally when component_add()
> fails so the probe error path matches the successful-probe remove path.
>
> Signed-off-by: Myeonghun Pak <mhun512@gmail.com>

Duplicate of [0]?

[0] https://lore.kernel.org/all/20260423200622.325076-1-osama.abdelkader@gmail.com/

Luca

--
Luca Ceresoli, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com