One option for the LVDS port of the LDB is to be connected to an
additional bridge, such as a LVDS to HDMI converter. Add support for
such case, along with the direct connection to a panel.

Signed-off-by: Francesco Valla <francesco@valla.it>
---
I was trying to add display support for the i.MX93 FRDM on top of the
patch sent some time ago by Fabian Pflug [1], using some of the work
already done by Alexander Stein but not yet merged [2], but then I
noticed that the support for LVDS-HDMI converter bridges was missing
from the LDB driver already present for the i.MX93.

Not a fail of the driver itself, obviously, but I wonder if/how the
existing i.MX8MP setups (e.g.: [3]), which use the same driver, work
correclty. Unfortunately I don't have the i.MX8MP hardware to test them.

Anyhow, a patch for such setup is attached; it was tested on the i.MX93
FRDM using [1] and [2] plus some more devicetree modifications.

[1] https://lore.kernel.org/all/20251022-fpg-nxp-imx93-frdm-v3-1-03ec40a1ccc0@pengutronix.de
[2] https://lore.kernel.org/all/20250304154929.1785200-1-alexander.stein@ew.tq-group.com
[3] https://elixir.bootlin.com/linux/v6.17.5/source/arch/arm64/boot/dts/freescale/imx8mp-evk-lvds0-imx-dlvds-hdmi-channel0.dtso

Regards,
Francesco
---
 drivers/gpu/drm/bridge/fsl-ldb.c | 26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/bridge/fsl-ldb.c b/drivers/gpu/drm/bridge/fsl-ldb.c
index 5c3cf37200bcee1db285c97e2b463c9355ee6acb..fad436f2e0bfac8b42096a6fcd0022da0f35284e 100644
--- a/drivers/gpu/drm/bridge/fsl-ldb.c
+++ b/drivers/gpu/drm/bridge/fsl-ldb.c
@@ -294,7 +294,6 @@ static int fsl_ldb_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct device_node *panel_node;
 	struct device_node *remote1, *remote2;
-	struct drm_panel *panel;
 	struct fsl_ldb *fsl_ldb;
 	int dual_link;
 
@@ -335,15 +334,24 @@ static int fsl_ldb_probe(struct platform_device *pdev)
 		fsl_ldb_is_dual(fsl_ldb) ? "dual-link mode" :
 		fsl_ldb->ch0_enabled ? "channel 0" : "channel 1");
 
-	panel = of_drm_find_panel(panel_node);
-	of_node_put(panel_node);
-	if (IS_ERR(panel))
-		return PTR_ERR(panel);
-
-	fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
-	if (IS_ERR(fsl_ldb->panel_bridge))
-		return PTR_ERR(fsl_ldb->panel_bridge);
+	/* First try to get an additional bridge, if not found go for a panel */
+	fsl_ldb->panel_bridge = of_drm_find_bridge(panel_node);
+	if (fsl_ldb->panel_bridge) {
+		of_node_put(panel_node);
+	} else {
+		struct drm_panel *panel;
 
+		panel = of_drm_find_panel(panel_node);
+		of_node_put(panel_node);
+		if (IS_ERR(panel))
+			return dev_err_probe(dev, PTR_ERR(panel),
+					     "Failed to find panel");
+
+		fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
+		if (IS_ERR(fsl_ldb->panel_bridge))
+			return dev_err_probe(dev, PTR_ERR(fsl_ldb->panel_bridge),
+					     "Failed to add panel bridge");
+	}
 
 	if (fsl_ldb_is_dual(fsl_ldb)) {
 		struct device_node *port1, *port2;

---
base-commit: fd57572253bc356330dbe5b233c2e1d8426c66fd
change-id: 20251028-imx93_ldb_bridge-3c011e7856dc

Best regards,
-- 
Francesco Valla <francesco@valla.it>

On 10/28/2025, Francesco Valla wrote:
> [You don't often get email from francesco@valla.it. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
> 
> One option for the LVDS port of the LDB is to be connected to an
> additional bridge, such as a LVDS to HDMI converter. Add support for
> such case, along with the direct connection to a panel.
> 
> Signed-off-by: Francesco Valla <francesco@valla.it>
> ---
> I was trying to add display support for the i.MX93 FRDM on top of the
> patch sent some time ago by Fabian Pflug [1], using some of the work
> already done by Alexander Stein but not yet merged [2], but then I
> noticed that the support for LVDS-HDMI converter bridges was missing
> from the LDB driver already present for the i.MX93.
> 
> Not a fail of the driver itself, obviously, but I wonder if/how the
> existing i.MX8MP setups (e.g.: [3]), which use the same driver, work
> correclty. Unfortunately I don't have the i.MX8MP hardware to test them.

[3] was in my previous patch series[a].  Only patch 6&7 of [a] are applied,
so for now [3] doesn't actually work with i.MX8MP.

And, patch 3 of [a] supports the external bridge this patch tries to support.

[b] is another patch series which includes my patch.

[a] https://patchwork.freedesktop.org/series/139266/#rev7
[b] https://patchwork.freedesktop.org/series/154381/

> 
> Anyhow, a patch for such setup is attached; it was tested on the i.MX93
> FRDM using [1] and [2] plus some more devicetree modifications.
> 
> [1] https://lore.kernel.org/all/20251022-fpg-nxp-imx93-frdm-v3-1-03ec40a1ccc0@pengutronix.de
> [2] https://lore.kernel.org/all/20250304154929.1785200-1-alexander.stein@ew.tq-group.com
> [3] https://elixir.bootlin.com/linux/v6.17.5/source/arch/arm64/boot/dts/freescale/imx8mp-evk-lvds0-imx-dlvds-hdmi-channel0.dtso
> 
> Regards,
> Francesco
> ---
>  drivers/gpu/drm/bridge/fsl-ldb.c | 26 +++++++++++++++++---------
>  1 file changed, 17 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/gpu/drm/bridge/fsl-ldb.c b/drivers/gpu/drm/bridge/fsl-ldb.c
> index 5c3cf37200bcee1db285c97e2b463c9355ee6acb..fad436f2e0bfac8b42096a6fcd0022da0f35284e 100644
> --- a/drivers/gpu/drm/bridge/fsl-ldb.c
> +++ b/drivers/gpu/drm/bridge/fsl-ldb.c
> @@ -294,7 +294,6 @@ static int fsl_ldb_probe(struct platform_device *pdev)
>         struct device *dev = &pdev->dev;
>         struct device_node *panel_node;
>         struct device_node *remote1, *remote2;
> -       struct drm_panel *panel;
>         struct fsl_ldb *fsl_ldb;
>         int dual_link;
> 
> @@ -335,15 +334,24 @@ static int fsl_ldb_probe(struct platform_device *pdev)
>                 fsl_ldb_is_dual(fsl_ldb) ? "dual-link mode" :
>                 fsl_ldb->ch0_enabled ? "channel 0" : "channel 1");
> 
> -       panel = of_drm_find_panel(panel_node);
> -       of_node_put(panel_node);
> -       if (IS_ERR(panel))
> -               return PTR_ERR(panel);
> -
> -       fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
> -       if (IS_ERR(fsl_ldb->panel_bridge))
> -               return PTR_ERR(fsl_ldb->panel_bridge);
> +       /* First try to get an additional bridge, if not found go for a panel */
> +       fsl_ldb->panel_bridge = of_drm_find_bridge(panel_node);
> +       if (fsl_ldb->panel_bridge) {
> +               of_node_put(panel_node);
> +       } else {
> +               struct drm_panel *panel;
> 
> +               panel = of_drm_find_panel(panel_node);
> +               of_node_put(panel_node);
> +               if (IS_ERR(panel))
> +                       return dev_err_probe(dev, PTR_ERR(panel),
> +                                            "Failed to find panel");
> +
> +               fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
> +               if (IS_ERR(fsl_ldb->panel_bridge))
> +                       return dev_err_probe(dev, PTR_ERR(fsl_ldb->panel_bridge),
> +                                            "Failed to add panel bridge");
> +       }
> 
>         if (fsl_ldb_is_dual(fsl_ldb)) {
>                 struct device_node *port1, *port2;
> 
> ---
> base-commit: fd57572253bc356330dbe5b233c2e1d8426c66fd
> change-id: 20251028-imx93_ldb_bridge-3c011e7856dc
> 
> Best regards,
> --
> Francesco Valla <francesco@valla.it>
> 


-- 
Regards,
Liu Ying

Hi,

On Wednesday, 29 October 2025 at 04:41:51 Liu Ying <victor.liu@nxp.com> wrote:
> On 10/28/2025, Francesco Valla wrote:
> > I was trying to add display support for the i.MX93 FRDM on top of the
> > patch sent some time ago by Fabian Pflug [1], using some of the work
> > already done by Alexander Stein but not yet merged [2], but then I
> > noticed that the support for LVDS-HDMI converter bridges was missing
> > from the LDB driver already present for the i.MX93.
> > 
> > Not a fail of the driver itself, obviously, but I wonder if/how the
> > existing i.MX8MP setups (e.g.: [3]), which use the same driver, work
> > correclty. Unfortunately I don't have the i.MX8MP hardware to test them.
> 
> [3] was in my previous patch series[a].  Only patch 6&7 of [a] are applied,
> so for now [3] doesn't actually work with i.MX8MP.
> 
> And, patch 3 of [a] supports the external bridge this patch tries to support.
> 
> [b] is another patch series which includes my patch.
> 
> [a] https://patchwork.freedesktop.org/series/139266/#rev7
> [b] https://patchwork.freedesktop.org/series/154381/
> 

Thank you for your answer, it's much clearer now.

I'll drop this one and wait for [b], then.


BR,
Francesco

Hi Francesco,

Am Dienstag, 28. Oktober 2025, 13:12:29 CET schrieb Francesco Valla:
> One option for the LVDS port of the LDB is to be connected to an
> additional bridge, such as a LVDS to HDMI converter. Add support for
> such case, along with the direct connection to a panel.
> 
> Signed-off-by: Francesco Valla <francesco@valla.it>
> ---
> I was trying to add display support for the i.MX93 FRDM on top of the
> patch sent some time ago by Fabian Pflug [1], using some of the work
> already done by Alexander Stein but not yet merged [2], but then I
> noticed that the support for LVDS-HDMI converter bridges was missing
> from the LDB driver already present for the i.MX93.
> 
> Not a fail of the driver itself, obviously, but I wonder if/how the
> existing i.MX8MP setups (e.g.: [3]), which use the same driver, work
> correclty. Unfortunately I don't have the i.MX8MP hardware to test them.
> 
> Anyhow, a patch for such setup is attached; it was tested on the i.MX93
> FRDM using [1] and [2] plus some more devicetree modifications.
> 
> [1] https://lore.kernel.org/all/20251022-fpg-nxp-imx93-frdm-v3-1-03ec40a1ccc0@pengutronix.de
> [2] https://lore.kernel.org/all/20250304154929.1785200-1-alexander.stein@ew.tq-group.com
> [3] https://elixir.bootlin.com/linux/v6.17.5/source/arch/arm64/boot/dts/freescale/imx8mp-evk-lvds0-imx-dlvds-hdmi-channel0.dtso
> 
> Regards,
> Francesco
> ---
>  drivers/gpu/drm/bridge/fsl-ldb.c | 26 +++++++++++++++++---------
>  1 file changed, 17 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/gpu/drm/bridge/fsl-ldb.c b/drivers/gpu/drm/bridge/fsl-ldb.c
> index 5c3cf37200bcee1db285c97e2b463c9355ee6acb..fad436f2e0bfac8b42096a6fcd0022da0f35284e 100644
> --- a/drivers/gpu/drm/bridge/fsl-ldb.c
> +++ b/drivers/gpu/drm/bridge/fsl-ldb.c
> @@ -294,7 +294,6 @@ static int fsl_ldb_probe(struct platform_device *pdev)
>  	struct device *dev = &pdev->dev;
>  	struct device_node *panel_node;
>  	struct device_node *remote1, *remote2;
> -	struct drm_panel *panel;
>  	struct fsl_ldb *fsl_ldb;
>  	int dual_link;
>  
> @@ -335,15 +334,24 @@ static int fsl_ldb_probe(struct platform_device *pdev)
>  		fsl_ldb_is_dual(fsl_ldb) ? "dual-link mode" :
>  		fsl_ldb->ch0_enabled ? "channel 0" : "channel 1");
>  
> -	panel = of_drm_find_panel(panel_node);
> -	of_node_put(panel_node);
> -	if (IS_ERR(panel))
> -		return PTR_ERR(panel);
> -
> -	fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
> -	if (IS_ERR(fsl_ldb->panel_bridge))
> -		return PTR_ERR(fsl_ldb->panel_bridge);
> +	/* First try to get an additional bridge, if not found go for a panel */
> +	fsl_ldb->panel_bridge = of_drm_find_bridge(panel_node);
> +	if (fsl_ldb->panel_bridge) {
> +		of_node_put(panel_node);
> +	} else {
> +		struct drm_panel *panel;
>  
> +		panel = of_drm_find_panel(panel_node);
> +		of_node_put(panel_node);
> +		if (IS_ERR(panel))
> +			return dev_err_probe(dev, PTR_ERR(panel),
> +					     "Failed to find panel");
> +
> +		fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
> +		if (IS_ERR(fsl_ldb->panel_bridge))
> +			return dev_err_probe(dev, PTR_ERR(fsl_ldb->panel_bridge),
> +					     "Failed to add panel bridge");
> +	}

Without looking into the details this somehow looks similar to
drm_of_find_panel_or_bridge(), or drmm_of_get_bridge for the managed variant.

Best regards,
Alexander

>  
>  	if (fsl_ldb_is_dual(fsl_ldb)) {
>  		struct device_node *port1, *port2;
> 
> ---
> base-commit: fd57572253bc356330dbe5b233c2e1d8426c66fd
> change-id: 20251028-imx93_ldb_bridge-3c011e7856dc
> 
> Best regards,
> 


-- 
TQ-Systems GmbH | Mühlstraße 2, Gut Delling | 82229 Seefeld, Germany
Amtsgericht München, HRB 105018
Geschäftsführer: Detlef Schneider, Rüdiger Stahl, Stefan Schneider
http://www.tq-group.com/

Hi,

On Tuesday, 28 October 2025 at 13:27:37 Alexander Stein <alexander.stein@ew.tq-group.com> wrote:
> Hi Francesco,
> 
> Am Dienstag, 28. Oktober 2025, 13:12:29 CET schrieb Francesco Valla:
> > One option for the LVDS port of the LDB is to be connected to an
> > additional bridge, such as a LVDS to HDMI converter. Add support for
> > such case, along with the direct connection to a panel.
> > 
> > Signed-off-by: Francesco Valla <francesco@valla.it>
> > ---
> > I was trying to add display support for the i.MX93 FRDM on top of the
> > patch sent some time ago by Fabian Pflug [1], using some of the work
> > already done by Alexander Stein but not yet merged [2], but then I
> > noticed that the support for LVDS-HDMI converter bridges was missing
> > from the LDB driver already present for the i.MX93.
> > 
> > Not a fail of the driver itself, obviously, but I wonder if/how the
> > existing i.MX8MP setups (e.g.: [3]), which use the same driver, work
> > correclty. Unfortunately I don't have the i.MX8MP hardware to test them.
> > 
> > Anyhow, a patch for such setup is attached; it was tested on the i.MX93
> > FRDM using [1] and [2] plus some more devicetree modifications.
> > 
> > [1] https://lore.kernel.org/all/20251022-fpg-nxp-imx93-frdm-v3-1-03ec40a1ccc0@pengutronix.de
> > [2] https://lore.kernel.org/all/20250304154929.1785200-1-alexander.stein@ew.tq-group.com
> > [3] https://elixir.bootlin.com/linux/v6.17.5/source/arch/arm64/boot/dts/freescale/imx8mp-evk-lvds0-imx-dlvds-hdmi-channel0.dtso
> > 
> > Regards,
> > Francesco
> > ---
> >  drivers/gpu/drm/bridge/fsl-ldb.c | 26 +++++++++++++++++---------
> >  1 file changed, 17 insertions(+), 9 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/bridge/fsl-ldb.c b/drivers/gpu/drm/bridge/fsl-ldb.c
> > index 5c3cf37200bcee1db285c97e2b463c9355ee6acb..fad436f2e0bfac8b42096a6fcd0022da0f35284e 100644
> > --- a/drivers/gpu/drm/bridge/fsl-ldb.c
> > +++ b/drivers/gpu/drm/bridge/fsl-ldb.c
> > @@ -294,7 +294,6 @@ static int fsl_ldb_probe(struct platform_device *pdev)
> >  	struct device *dev = &pdev->dev;
> >  	struct device_node *panel_node;
> >  	struct device_node *remote1, *remote2;
> > -	struct drm_panel *panel;
> >  	struct fsl_ldb *fsl_ldb;
> >  	int dual_link;
> >  
> > @@ -335,15 +334,24 @@ static int fsl_ldb_probe(struct platform_device *pdev)
> >  		fsl_ldb_is_dual(fsl_ldb) ? "dual-link mode" :
> >  		fsl_ldb->ch0_enabled ? "channel 0" : "channel 1");
> >  
> > -	panel = of_drm_find_panel(panel_node);
> > -	of_node_put(panel_node);
> > -	if (IS_ERR(panel))
> > -		return PTR_ERR(panel);
> > -
> > -	fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
> > -	if (IS_ERR(fsl_ldb->panel_bridge))
> > -		return PTR_ERR(fsl_ldb->panel_bridge);
> > +	/* First try to get an additional bridge, if not found go for a panel */
> > +	fsl_ldb->panel_bridge = of_drm_find_bridge(panel_node);
> > +	if (fsl_ldb->panel_bridge) {
> > +		of_node_put(panel_node);
> > +	} else {
> > +		struct drm_panel *panel;
> >  
> > +		panel = of_drm_find_panel(panel_node);
> > +		of_node_put(panel_node);
> > +		if (IS_ERR(panel))
> > +			return dev_err_probe(dev, PTR_ERR(panel),
> > +					     "Failed to find panel");
> > +
> > +		fsl_ldb->panel_bridge = devm_drm_panel_bridge_add(dev, panel);
> > +		if (IS_ERR(fsl_ldb->panel_bridge))
> > +			return dev_err_probe(dev, PTR_ERR(fsl_ldb->panel_bridge),
> > +					     "Failed to add panel bridge");
> > +	}
> 
> Without looking into the details this somehow looks similar to
> drm_of_find_panel_or_bridge(), or drmm_of_get_bridge for the managed variant.
> 
> Best regards,
> Alexander
> 
> >  
> >  	if (fsl_ldb_is_dual(fsl_ldb)) {
> >  		struct device_node *port1, *port2;
> > 
> > ---
> > base-commit: fd57572253bc356330dbe5b233c2e1d8426c66fd
> > change-id: 20251028-imx93_ldb_bridge-3c011e7856dc
> > 
> > Best regards,
> > 
> 
> 
> 

Well, I did not know of devm_drm_of_get_bridge(), and unfortunately the
driver I checked (Samsung DSIM) has not yet been updated to use it.
Sorry for the noise.

I'll wait for more feedback (if any) and then send a v2 - which looks
much cleaner now.

Thank you!

Regards,
Francesco