Re: [PATCH] drm/bridge: lontium-lt9611uxc: use irq as optional parameter

[Luca Ceresoli]

Hello,

On Mon Dec 8, 2025 at 5:17 PM CET, Vladimir Yakovlev wrote:
> On some systems the interrupt pin may not be used.
> In this case we exclude DRM_BRIDGE_OP_HPD from supported operations,
> after which a polling thread is started to detect the connection.
> (the default polling period for DRM is 10 seconds)
>
> Signed-off-by: Vladimir Yakovlev <vovchkir@xxxxxxxxx>
> ---
>  drivers/gpu/drm/bridge/lontium-lt9611uxc.c | 26 ++++++++++++++--------
>  1 file changed, 17 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/gpu/drm/bridge/lontium-lt9611uxc.c b/drivers/gpu/drm/bridge/lontium-lt9611uxc.c
> index 3868ad05e011..4d550cd1a9a1 100644
> --- a/drivers/gpu/drm/bridge/lontium-lt9611uxc.c
> +++ b/drivers/gpu/drm/bridge/lontium-lt9611uxc.c
> @@ -851,19 +851,25 @@ static int lt9611uxc_probe(struct i2c_client *client)
>  	init_waitqueue_head(&lt9611uxc->wq);
>  	INIT_WORK(&lt9611uxc->work, lt9611uxc_hpd_work);
>
> -	ret = request_threaded_irq(client->irq, NULL,
> -				   lt9611uxc_irq_thread_handler,
> -				   IRQF_ONESHOT, "lt9611uxc", lt9611uxc);
> -	if (ret) {
> -		dev_err(dev, "failed to request irq\n");
> -		goto err_disable_regulators;
> +	if (client->irq) {
> +		ret = request_threaded_irq(client->irq, NULL,
> +					   lt9611uxc_irq_thread_handler,
> +					   IRQF_ONESHOT, "lt9611uxc", lt9611uxc);

This is per se correct, but while you're touching this it's a good
opportunity to switch to devm_request_threaded_irq(). This would simplify
the removal code too (see below).

> +		if (ret) {
> +			dev_err(dev, "failed to request irq\n");
> +			goto err_disable_regulators;
> +		}
> +		dev_dbg(dev, "Uses IRQ\n");
> +	} else {
> +		dev_warn(dev, "The interrupt (IRQ) is not specified in the DTS.\n");
> +		dev_warn(dev, "Check the interrupt (IRQ) or polling will be used!!!\n");

If hardware without this IRQ is supported, there is no reason to warn. You
can have a dev_dbg() at most, or even nothing at all.

>  	}
>
>  	i2c_set_clientdata(client, lt9611uxc);
>
>  	lt9611uxc->bridge.of_node = client->dev.of_node;
>  	lt9611uxc->bridge.ops = DRM_BRIDGE_OP_DETECT | DRM_BRIDGE_OP_EDID;
> -	if (lt9611uxc->hpd_supported)
> +	if (lt9611uxc->hpd_supported && client->irq)
>  		lt9611uxc->bridge.ops |= DRM_BRIDGE_OP_HPD;
>  	lt9611uxc->bridge.type = DRM_MODE_CONNECTOR_HDMIA;
>
> @@ -892,7 +898,8 @@ static int lt9611uxc_probe(struct i2c_client *client)
>  	return 0;
>
>  err_remove_bridge:
> -	free_irq(client->irq, lt9611uxc);
> +	if (client->irq)
> +		free_irq(client->irq, lt9611uxc);

With devm_request_threaded_irq() you can remove entirely the free_irq()
here...

>  	cancel_work_sync(&lt9611uxc->work);
>  	drm_bridge_remove(&lt9611uxc->bridge);
>
> @@ -910,7 +917,8 @@ static void lt9611uxc_remove(struct i2c_client *client)
>  {
>  	struct lt9611uxc *lt9611uxc = i2c_get_clientdata(client);
>
> -	free_irq(client->irq, lt9611uxc);
> +	if (client->irq)
> +		free_irq(client->irq, lt9611uxc);

...as well as here.

So I'd suggest to add a new patch to switch to devm_request_threaded_irq(),
followed by this one, which will be a simpler.

Best regards,
Luca

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