[PATCH v3 0/3] drm: remove deprecated *_of_get_bridge() from non-OF drivers

[PATCH v3 0/3] drm: remove deprecated *_of_get_bridge() from non-OF drivers

[Luca Ceresoli]

Replace the deprecated *_of_get_bridge() calls with
of_drm_get_bridge_by_endpoint() for all drivers which be built without
CONFIG_OF.

Some of those drivers had a build failure for configs without
CONFIG_OF. This series used have fixes for those failures plus other
patches to convert all other such drivers to
of_drm_get_bridge_by_endpoint().

All the fixes are already applied, so since v3 there are only the
improvement patches not yet applied.

Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
---
Changes in v3:
- Patch 1: re-add comment, updated
- Patch 2: fix bridge reference leak in error path
- Patch 3: new patch suggested during v2 review
- Link to v2: https://patch.msgid.link/20260929-drm-fix-of_get_bridge-build-regression-v2-0-6da431d8f180@bootlin.com

Changes in v2:
- Fix bugs in patches 2,3,4
- Fix removed function name in patches 5,6
- Mention lkp reports in patches 2,3
- Add review tags on patches not modified since v1
- Link to v1: https://patch.msgid.link/20260929-drm-fix-of_get_bridge-build-regression-v1-0-ccb7e602daad@bootlin.com

---
Luca Ceresoli (3):
      drm/vc4: dpi: stop using deprecated drmm_of_get_bridge()
      drm/vc4: dsi: stop using deprecated drmm_of_get_bridge()
      drm/vc4: dpi: inline vc4_dpi_init_bridge() into vc4_dpi_bind()

 drivers/gpu/drm/vc4/vc4_dpi.c | 41 ++++++++++++++---------------------------
 drivers/gpu/drm/vc4/vc4_dsi.c | 17 ++++++++++-------
 2 files changed, 24 insertions(+), 34 deletions(-)
---
base-commit: c1e2ea83145afeb5a693cff3b32c0e52b0608649
change-id: 20260929-drm-fix-of_get_bridge-build-regression-98f22f8a1dcc

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

[PATCH v3 1/3] drm/vc4: dpi: stop using deprecated drmm_of_get_bridge()

[Luca Ceresoli]

drmm_of_get_bridge() is deprecated, switch to
of_drm_get_bridge_by_endpoint(). As of_drm_get_bridge_by_endpoint() returns
a refcounted bridge, ensure it is put.

drm_panel.h becomes unused, drop the include.

Also remove a comment that is now obvious: now panels are just bridges.

Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>

---

Changes in v3:
- re-add and update comment
---
 drivers/gpu/drm/vc4/vc4_dpi.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c
index 53f36626a50d..feefefcf3ad7 100644
--- a/drivers/gpu/drm/vc4/vc4_dpi.c
+++ b/drivers/gpu/drm/vc4/vc4_dpi.c
@@ -16,7 +16,6 @@
 #include <drm/drm_drv.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_of.h>
-#include <drm/drm_panel.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_simple_kms_helper.h>
@@ -279,16 +278,13 @@ static const struct of_device_id vc4_dpi_dt_match[] = {
 	{}
 };
 
-/* Sets up the next link in the display chain, whether it's a panel or
- * a bridge.
- */
+/* Sets up the next link in the display chain */
 static int vc4_dpi_init_bridge(struct vc4_dpi *dpi)
 {
-	struct drm_device *drm = dpi->encoder.base.dev;
 	struct device *dev = &dpi->pdev->dev;
-	struct drm_bridge *bridge;
 
-	bridge = drmm_of_get_bridge(drm, dev->of_node, 0, 0);
+	struct drm_bridge *bridge __free(drm_bridge_put) =
+		of_drm_get_bridge_by_endpoint(dev->of_node, 0, 0);
 	if (IS_ERR(bridge)) {
 		/* If nothing was connected in the DT, that's not an
 		 * error.

-- 
2.56.0

[PATCH v3 2/3] drm/vc4: dsi: stop using deprecated drmm_of_get_bridge()

[Luca Ceresoli]

drmm_of_get_bridge() is deprecated, switch to
of_drm_get_bridge_by_endpoint(). As of_drm_get_bridge_by_endpoint() returns
a refcounted bridge, ensure it is put by using bridge.next_bridge.

drm_panel.h becomes unused, drop the include.

Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>

---

Changes in v3:
- fix bridge reference leak in error path
---
 drivers/gpu/drm/vc4/vc4_dsi.c | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c
index 0148f2befed1..675bd0fc1391 100644
--- a/drivers/gpu/drm/vc4/vc4_dsi.c
+++ b/drivers/gpu/drm/vc4/vc4_dsi.c
@@ -35,7 +35,6 @@
 #include <drm/drm_edid.h>
 #include <drm/drm_mipi_dsi.h>
 #include <drm/drm_of.h>
-#include <drm/drm_panel.h>
 #include <drm/drm_print.h>
 #include <drm/drm_probe_helper.h>
 #include <drm/drm_simple_kms_helper.h>
@@ -555,7 +554,6 @@ struct vc4_dsi {
 
 	struct platform_device *pdev;
 
-	struct drm_bridge *out_bridge;
 	struct drm_bridge bridge;
 
 	void __iomem *regs;
@@ -1165,7 +1163,7 @@ static int vc4_dsi_bridge_attach(struct drm_bridge *bridge,
 	struct vc4_dsi *dsi = bridge_to_vc4_dsi(bridge);
 
 	/* Attach the panel or bridge to the dsi bridge */
-	return drm_bridge_attach(encoder, dsi->out_bridge,
+	return drm_bridge_attach(encoder, dsi->bridge.next_bridge,
 				 &dsi->bridge, flags);
 }
 
@@ -1749,9 +1747,10 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
 		return ret;
 	}
 
-	dsi->out_bridge = drmm_of_get_bridge(drm, dev->of_node, 0, 0);
-	if (IS_ERR(dsi->out_bridge))
-		return PTR_ERR(dsi->out_bridge);
+	struct drm_bridge *next_bridge __free(drm_bridge_put) =
+		of_drm_get_bridge_by_endpoint(dev->of_node, 0, 0);
+	if (IS_ERR(next_bridge))
+		return PTR_ERR(next_bridge);
 
 	/* The esc clock rate is supposed to always be 100Mhz. */
 	ret = clk_set_rate(dsi->escape_clock, 100 * 1000000);
@@ -1775,9 +1774,13 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
 	if (ret)
 		return ret;
 
+	dsi->bridge.next_bridge = drm_bridge_get(next_bridge);
+
 	ret = drm_bridge_attach(encoder, &dsi->bridge, NULL, 0);
-	if (ret)
+	if (ret) {
+		drm_bridge_clear_and_put(&dsi->bridge.next_bridge);
 		return ret;
+	}
 
 	return 0;
 }

-- 
2.56.0

[PATCH v3 3/3] drm/vc4: dpi: inline vc4_dpi_init_bridge() into vc4_dpi_bind()

[Luca Ceresoli]

There isn't much benefit in this small static function. Inline it in the
caller.

Link: https://lore.kernel.org/lkml/accaa9f4ad05ef73071e5cfb5368fcbdd394d3ab.camel@bootlin.com/
Suggested-by: Paul Cercueil <paul.cercueil@bootlin.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>

---

New patch in v3
---
 drivers/gpu/drm/vc4/vc4_dpi.c | 37 ++++++++++++++-----------------------
 1 file changed, 14 insertions(+), 23 deletions(-)

diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c
index feefefcf3ad7..5439e73da0d4 100644
--- a/drivers/gpu/drm/vc4/vc4_dpi.c
+++ b/drivers/gpu/drm/vc4/vc4_dpi.c
@@ -278,26 +278,6 @@ static const struct of_device_id vc4_dpi_dt_match[] = {
 	{}
 };
 
-/* Sets up the next link in the display chain */
-static int vc4_dpi_init_bridge(struct vc4_dpi *dpi)
-{
-	struct device *dev = &dpi->pdev->dev;
-
-	struct drm_bridge *bridge __free(drm_bridge_put) =
-		of_drm_get_bridge_by_endpoint(dev->of_node, 0, 0);
-	if (IS_ERR(bridge)) {
-		/* If nothing was connected in the DT, that's not an
-		 * error.
-		 */
-		if (PTR_ERR(bridge) == -ENODEV)
-			return 0;
-		else
-			return PTR_ERR(bridge);
-	}
-
-	return drm_bridge_attach(&dpi->encoder.base, bridge, NULL, 0);
-}
-
 static void vc4_dpi_disable_clock(void *ptr)
 {
 	struct vc4_dpi *dpi = ptr;
@@ -366,9 +346,20 @@ static int vc4_dpi_bind(struct device *dev, struct device *master, void *data)
 
 	drm_encoder_helper_add(&dpi->encoder.base, &vc4_dpi_encoder_helper_funcs);
 
-	ret = vc4_dpi_init_bridge(dpi);
-	if (ret)
-		return ret;
+	/* Set up the next link in the display chain */
+	struct drm_bridge *bridge __free(drm_bridge_put) =
+		of_drm_get_bridge_by_endpoint(dev->of_node, 0, 0);
+	if (IS_ERR(bridge)) {
+		/* If nothing was connected in the DT (-ENODEV), that's not
+		 * an error.
+		 */
+		if (PTR_ERR(bridge) != -ENODEV)
+			return PTR_ERR(bridge);
+	} else {
+		ret = drm_bridge_attach(&dpi->encoder.base, bridge, NULL, 0);
+		if (ret)
+			return ret;
+	}
 
 	dev_set_drvdata(dev, dpi);
 

-- 
2.56.0

Re: [PATCH v3 2/3] drm/vc4: dsi: stop using deprecated drmm_of_get_bridge()

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] drm/vc4: reference leak of next_bridge on component unbind/rebind
--

commit 01a94378aeae8cb42a9f68ed6384a34ccaa162a0
Author: Luca Ceresoli <luca.ceresoli@bootlin.com>

drm/vc4: dsi: stop using deprecated drmm_of_get_bridge()

drmm_of_get_bridge() is deprecated, switch to
of_drm_get_bridge_by_endpoint(). As of_drm_get_bridge_by_endpoint() returns
a refcounted bridge, ensure it is put by using bridge.next_bridge.

> diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c
> index 0148f2befed14..675bd0fc13913 100644
> --- a/drivers/gpu/drm/vc4/vc4_dsi.c
> +++ b/drivers/gpu/drm/vc4/vc4_dsi.c
[ ... ]
> @@ -1775,9 +1774,13 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
>  	if (ret)
>  		return ret;
>  
> +	dsi->bridge.next_bridge = drm_bridge_get(next_bridge);
> +

[Severity: Medium]
Does this code introduce a regression that leaks the bridge reference across
component unbind/rebind cycles?

During component unbind, vc4_dsi_release_action() only drops the reference
to dsi->bridge acquired during bind:

static void vc4_dsi_release_action(struct drm_device *drm, void *ptr)
{
	struct vc4_dsi *dsi = ptr;

	drm_bridge_put(&dsi->bridge);
}

Because dsi->bridge outlives the component unbind, __drm_bridge_free() is
not executed, and the reference to next_bridge is not automatically released.

When the component is subsequently rebound, vc4_dsi_bind() overwrites
dsi->bridge.next_bridge here without freeing the previous pointer,
permanently leaking the bridge reference. Does next_bridge need to be
explicitly cleared in the teardown path?

>  	ret = drm_bridge_attach(encoder, &dsi->bridge, NULL, 0);
> -	if (ret)
> +	if (ret) {
> +		drm_bridge_clear_and_put(&dsi->bridge.next_bridge);
>  		return ret;
> +	}
>  
>  	return 0;
>  }

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261013-drm-fix-of_get_bridge-build-regression-v3-0-0b4c0b65be82@bootlin.com?part=2

Re: [PATCH v3 2/3] drm/vc4: dsi: stop using deprecated drmm_of_get_bridge()

[Luca Ceresoli]

On Tue Oct 13, 2026 at 5:35 PM CEST, sashiko-bot wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] drm/vc4: reference leak of next_bridge on component unbind/rebind
> --
>
> commit 01a94378aeae8cb42a9f68ed6384a34ccaa162a0
> Author: Luca Ceresoli <luca.ceresoli@bootlin.com>
>
> drm/vc4: dsi: stop using deprecated drmm_of_get_bridge()
>
> drmm_of_get_bridge() is deprecated, switch to
> of_drm_get_bridge_by_endpoint(). As of_drm_get_bridge_by_endpoint() returns
> a refcounted bridge, ensure it is put by using bridge.next_bridge.
>
>> diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c
>> index 0148f2befed14..675bd0fc13913 100644
>> --- a/drivers/gpu/drm/vc4/vc4_dsi.c
>> +++ b/drivers/gpu/drm/vc4/vc4_dsi.c
> [ ... ]
>> @@ -1775,9 +1774,13 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
>>  	if (ret)
>>  		return ret;
>>
>> +	dsi->bridge.next_bridge = drm_bridge_get(next_bridge);
>> +
>
> [Severity: Medium]
> Does this code introduce a regression that leaks the bridge reference across
> component unbind/rebind cycles?
>
> During component unbind, vc4_dsi_release_action() only drops the reference
> to dsi->bridge acquired during bind:
>
> static void vc4_dsi_release_action(struct drm_device *drm, void *ptr)
> {
> 	struct vc4_dsi *dsi = ptr;
>
> 	drm_bridge_put(&dsi->bridge);
> }
>
> Because dsi->bridge outlives the component unbind, __drm_bridge_free() is
> not executed, and the reference to next_bridge is not automatically released.
>
> When the component is subsequently rebound, vc4_dsi_bind() overwrites
> dsi->bridge.next_bridge here without freeing the previous pointer,
> permanently leaking the bridge reference. Does next_bridge need to be
> explicitly cleared in the teardown path?

Ah, right. I'm fixing this in v4 by calling
drm_bridge_clear_and_put(&dsi->bridge.next_bridge).

Not sure adding it to vc4_dsi_release_action() would be correct in case of
unbind/rebind, so I'm probably adding a new vc4_dsi_unbind() for that.

Luca

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