From: Lee Jones <lee@kernel.org>
To: Yige Jiang <yigejiang86@gmail.com>
Cc: linux-leds@vger.kernel.org, Pavel Machek <pavel@kernel.org>,
Griffin Kroah-Hartman <griffin@kroah.com>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] leds: pwm-multicolor: Manage the multi-led node reference
Date: Fri, 13 Nov 2026 11:04:24 +0000 [thread overview]
Message-ID: <179370050695.793788.4199070054904484695@kernel.org> (raw)
In-Reply-To: <20261030020521.9399-1-yigejiang86@gmail.com>
On Fri, 30 Oct 2026, Yige Jiang wrote:
> led_pwm_mc_probe() leaks the multi-led node reference on successful
> probe and if the final PWM update fails. LED registration borrows
> the node, so it must remain valid until the LED is unregistered.
>
> Register a managed put action before LED registration. Reverse devres
> order then unregisters the LED before releasing the node on unbind
> or probe failure. Drop the manual error-path put and return directly
> on probe errors, leaving the reference cleanup to devres.
>
> Fixes: 9fa2762110dd ("leds: Add PWM multicolor driver")
> Assisted-by: LLM
I thought we dropped these now?
> Signed-off-by: Yige Jiang <yigejiang86@gmail.com>
> ---
> Changes in v2:
> - Capitalize the subject description to follow the LED subsystem style.
> - Remove the redundant release_mcnode label and return directly from
> the probe error paths.
>
> v1: https://lore.kernel.org/r/rc-review-20261029-ai27-27-2c77ed65cec2@gmail.com
>
> Based on linux-next next-20261016 (2efb7a11f1aa).
> Compile-tested: arm64, GCC 14.2.0, W=1, CONFIG_OF_DYNAMIC=y.
> COMPILE_TEST=y.
> The object was rebuilt from this exact base with v2 applied.
> No full kernel or module-link, hardware or runtime test was performed.
>
> drivers/leds/rgb/leds-pwm-multicolor.c | 31 +++++++++++++-------------
> 1 file changed, 16 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/leds/rgb/leds-pwm-multicolor.c b/drivers/leds/rgb/leds-pwm-multicolor.c
> index 6365459ec27f17187804153cde6cb1805eca33e9..10527a64d1982733a43b76bd8bc123aba95b6f72 100644
> --- a/drivers/leds/rgb/leds-pwm-multicolor.c
> +++ b/drivers/leds/rgb/leds-pwm-multicolor.c
> @@ -106,6 +106,11 @@ release_fwnode:
> return ret;
> }
>
> +static void led_pwm_mc_put_node(void *data)
> +{
> + fwnode_handle_put(data);
> +}
What about?
struct fwnode_handle *mcnode __free(fwnode_handle) =
device_get_named_child_node(&pdev->dev, "multi-led");
> +
> static int led_pwm_mc_probe(struct platform_device *pdev)
> {
> struct fwnode_handle *mcnode;
> @@ -121,22 +126,22 @@ static int led_pwm_mc_probe(struct platform_device *pdev)
> return dev_err_probe(&pdev->dev, -ENODEV,
> "expected multi-led node\n");
>
> + ret = devm_add_action_or_reset(&pdev->dev, led_pwm_mc_put_node, mcnode);
> + if (ret)
> + return ret;
> +
> /* count the nodes inside the multi-led node */
> count = fwnode_get_child_node_count(mcnode);
>
> priv = devm_kzalloc(&pdev->dev, struct_size(priv, leds, count),
> GFP_KERNEL);
> - if (!priv) {
> - ret = -ENOMEM;
> - goto release_mcnode;
> - }
> + if (!priv)
> + return -ENOMEM;
> mutex_init(&priv->lock);
>
> subled = devm_kcalloc(&pdev->dev, count, sizeof(*subled), GFP_KERNEL);
> - if (!subled) {
> - ret = -ENOMEM;
> - goto release_mcnode;
> - }
> + if (!subled)
> + return -ENOMEM;
> priv->mc_cdev.subled_info = subled;
>
> /* init the multicolor's LED class device */
> @@ -144,14 +149,14 @@ static int led_pwm_mc_probe(struct platform_device *pdev)
> ret = fwnode_property_read_u32(mcnode, "max-brightness",
> &cdev->max_brightness);
> if (ret)
> - goto release_mcnode;
> + return ret;
>
> cdev->flags = LED_CORE_SUSPENDRESUME;
> cdev->brightness_set_blocking = led_pwm_mc_set;
>
> ret = iterate_subleds(&pdev->dev, priv, mcnode);
> if (ret)
> - goto release_mcnode;
> + return ret;
>
> init_data.fwnode = mcnode;
> ret = devm_led_classdev_multicolor_register_ext(&pdev->dev,
> @@ -161,7 +166,7 @@ static int led_pwm_mc_probe(struct platform_device *pdev)
> dev_err(&pdev->dev,
> "failed to register multicolor PWM led for %s: %d\n",
> cdev->name, ret);
> - goto release_mcnode;
> + return ret;
> }
>
> ret = led_pwm_mc_set(cdev, cdev->brightness);
> @@ -172,10 +177,6 @@ static int led_pwm_mc_probe(struct platform_device *pdev)
>
> platform_set_drvdata(pdev, priv);
> return 0;
> -
> -release_mcnode:
> - fwnode_handle_put(mcnode);
> - return ret;
> }
>
> static const struct of_device_id of_pwm_leds_mc_match[] = {
>
> base-commit: 2efb7a11f1aa4d2dc285fe2680cca22cb9f72f64
>
--
Lee Jones
prev parent reply other threads:[~2026-11-13 11:04 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-30 2:05 Yige Jiang
2026-11-13 11:04 ` Lee Jones [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=179370050695.793788.4199070054904484695@kernel.org \
--to=lee@kernel.org \
--cc=griffin@kroah.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-leds@vger.kernel.org \
--cc=pavel@kernel.org \
--cc=yigejiang86@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®