linux-kernel - Re: [PATCH] platform/x86: int3472: Fix double free of GPIO device during unregister

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <5007d7f0-76ff-41fd-a371-05922c97f8ef@ideasonboard.com>
Date: Tue, 28 Oct 2025 08:55:07 +0000
From: Dan Scally <dan.scally@...asonboard.com>
To: Qiu Wenbo <qiuwenbo@...me.org>, Daniel Scally <djrscally@...il.com>,
 Hans de Goede <hansg@...nel.org>
Cc: Qiu Wenbo <qiuwenbo@...insec.com.cn>,
 platform-driver-x86@...r.kernel.org, linux-kernel@...r.kernel.org,
 Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
 Sakari Ailus <sakari.ailus@...ux.intel.com>,
 Andy Shevchenko <andy@...nel.org>
Subject: Re: [PATCH] platform/x86: int3472: Fix double free of GPIO device
 during unregister

Hi Qiu

On 24/10/2025 06:05, Qiu Wenbo wrote:
> From: Qiu Wenbo <qiuwenbo@...insec.com.cn>
> 
> regulator_unregister() already frees the associated GPIO device. On
> ThinkPad X9 (Lunar Lake), this causes a double free issue that leads to
> random failures when other drivers (typically Intel THC) attempt to
> allocate interrupts. The root cause is that the reference count of the
> pinctrl_intel_platform module unexpectedly drops to zero when this
> driver defers its probe.
> 
> This behavior can also be reproduced by unloading the module directly.
> 
> Fix the issue by removing the redundant release of the GPIO device
> during regulator unregistration.
> 
> Fixes: 1e5d088a52c2 ("platform/x86: int3472: Stop using devm_gpiod_get()")
> 
> Signed-off-by: Qiu Wenbo <qiuwenbo@...insec.com.cn>
> ---

The change looks good to me, so:

Reviewed-by: Daniel Scally <dan.scally@...asonboard.com>

However the Fixes tag I wonder about; devm_gpiod_get() will also result in a call to gpiod_put() 
when the module is unloaded; doesn't that mean that the same issue  will occur before that commit?

Thanks
Dan



>   drivers/platform/x86/intel/int3472/clk_and_regulator.c | 5 +----
>   include/linux/platform_data/x86/int3472.h              | 1 -
>   2 files changed, 1 insertion(+), 5 deletions(-)
> 
> diff --git a/drivers/platform/x86/intel/int3472/clk_and_regulator.c b/drivers/platform/x86/intel/int3472/clk_and_regulator.c
> index 476ec24d37020..9e052b164a1ab 100644
> --- a/drivers/platform/x86/intel/int3472/clk_and_regulator.c
> +++ b/drivers/platform/x86/intel/int3472/clk_and_regulator.c
> @@ -245,15 +245,12 @@ int skl_int3472_register_regulator(struct int3472_discrete_device *int3472,
>   	if (IS_ERR(regulator->rdev))
>   		return PTR_ERR(regulator->rdev);
>   
> -	int3472->regulators[int3472->n_regulator_gpios].ena_gpio = gpio;
>   	int3472->n_regulator_gpios++;
>   	return 0;
>   }
>   
>   void skl_int3472_unregister_regulator(struct int3472_discrete_device *int3472)
>   {
> -	for (int i = 0; i < int3472->n_regulator_gpios; i++) {
> +	for (int i = 0; i < int3472->n_regulator_gpios; i++)
>   		regulator_unregister(int3472->regulators[i].rdev);
> -		gpiod_put(int3472->regulators[i].ena_gpio);
> -	}
>   }
> diff --git a/include/linux/platform_data/x86/int3472.h b/include/linux/platform_data/x86/int3472.h
> index 1571e9157fa50..b1b837583d544 100644
> --- a/include/linux/platform_data/x86/int3472.h
> +++ b/include/linux/platform_data/x86/int3472.h
> @@ -100,7 +100,6 @@ struct int3472_gpio_regulator {
>   	struct regulator_consumer_supply supply_map[GPIO_REGULATOR_SUPPLY_MAP_COUNT * 2];
>   	char supply_name_upper[GPIO_SUPPLY_NAME_LENGTH];
>   	char regulator_name[GPIO_REGULATOR_NAME_LENGTH];
> -	struct gpio_desc *ena_gpio;
>   	struct regulator_dev *rdev;
>   	struct regulator_desc rdesc;
>   };