Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <ac28b69a-df20-ee17-a567-026096ed5498@samsung.com>
Date:   Fri, 29 Apr 2022 23:56:56 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0)
        Gecko/20100101 Thunderbird/91.8.0
Subject: Re: [PATCH 1/2] rpmsg: Fix calling device_lock() on non-initialized
 device
Content-Language: en-US
To:     Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>,
        Bjorn Andersson <bjorn.andersson@linaro.org>,
        Mathieu Poirier <mathieu.poirier@linaro.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        linux-remoteproc@vger.kernel.org, linux-kernel@vger.kernel.org
From:   Marek Szyprowski <m.szyprowski@samsung.com>
In-Reply-To: <20220429195946.1061725-2-krzysztof.kozlowski@linaro.org>
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="utf-8"
CMS-TYPE: 201P
References: <20220429195946.1061725-1-krzysztof.kozlowski@linaro.org>
        <CGME20220429195958eucas1p11ae281abca3e039a337cf46be15d8b86@eucas1p1.samsung.com>
        <20220429195946.1061725-2-krzysztof.kozlowski@linaro.org>
Precedence: bulk

On 29.04.2022 21:59, Krzysztof Kozlowski wrote:
> driver_set_override() helper uses device_lock() so it should not be
> called before rpmsg_register_device() (which calls device_register()).
> Effect can be seen with CONFIG_DEBUG_MUTEXES:
>
>    DEBUG_LOCKS_WARN_ON(lock->magic != lock)
>    WARNING: CPU: 3 PID: 57 at kernel/locking/mutex.c:582 __mutex_lock+0x1ec/0x430
>    ...
>    Call trace:
>     __mutex_lock+0x1ec/0x430
>     mutex_lock_nested+0x44/0x50
>     driver_set_override+0x124/0x150
>     qcom_glink_native_probe+0x30c/0x3b0
>     glink_rpm_probe+0x274/0x350
>     platform_probe+0x6c/0xe0
>     really_probe+0x17c/0x3d0
>     __driver_probe_device+0x114/0x190
>     driver_probe_device+0x3c/0xf0
>     ...
>
> Refactor the rpmsg_register_device() function to use two-step device
> registering (initialization + add) and call driver_set_override() in
> proper moment.
>
> This moves the code around, so while at it also NULL-ify the
> rpdev->driver_override in error path to be sure it won't be kfree()
> second time.
>
> Fixes: 42cd402b8fd4 ("rpmsg: Fix kfree() of static memory on setting driver_override")
> Reported-by: Marek Szyprowski <m.szyprowski@samsung.com>
> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
>
> Commit SHA from linux-next - Greg's tree.
> ---
>   drivers/rpmsg/rpmsg_core.c     | 33 ++++++++++++++++++++++++++++++---
>   drivers/rpmsg/rpmsg_internal.h | 14 +-------------
>   drivers/rpmsg/rpmsg_ns.c       | 14 +-------------
>   include/linux/rpmsg.h          |  8 ++++++++
>   4 files changed, 40 insertions(+), 29 deletions(-)
>
> diff --git a/drivers/rpmsg/rpmsg_core.c b/drivers/rpmsg/rpmsg_core.c
> index 95fc283f6af7..4938fc4eff00 100644
> --- a/drivers/rpmsg/rpmsg_core.c
> +++ b/drivers/rpmsg/rpmsg_core.c
> @@ -593,24 +593,51 @@ static struct bus_type rpmsg_bus = {
>   	.remove		= rpmsg_dev_remove,
>   };
>   
> -int rpmsg_register_device(struct rpmsg_device *rpdev)
> +/*
> + * A helper for registering rpmsg device with driver override and name.
> + * Drivers should not be using it, but instead rpmsg_register_device().
> + */
> +int rpmsg_register_device_override(struct rpmsg_device *rpdev,
> +				   const char *driver_override)
>   {
>   	struct device *dev = &rpdev->dev;
>   	int ret;
>   
> +	if (driver_override)
> +		strcpy(rpdev->id.name, driver_override);
> +
>   	dev_set_name(&rpdev->dev, "%s.%s.%d.%d", dev_name(dev->parent),
>   		     rpdev->id.name, rpdev->src, rpdev->dst);
>   
>   	rpdev->dev.bus = &rpmsg_bus;
>   
> -	ret = device_register(&rpdev->dev);
> +	device_initialize(dev);
> +	if (driver_override) {
> +		ret = driver_set_override(dev, &rpdev->driver_override,
> +					  driver_override,
> +					  strlen(driver_override));
> +		if (ret) {
> +			dev_err(dev, "device_set_override failed: %d\n", ret);
> +			return ret;
> +		}
> +	}
> +
> +	ret = device_add(dev);
>   	if (ret) {
> -		dev_err(dev, "device_register failed: %d\n", ret);
> +		dev_err(dev, "device_add failed: %d\n", ret);
> +		kfree(rpdev->driver_override);
> +		rpdev->driver_override = NULL;
>   		put_device(&rpdev->dev);
>   	}
>   
>   	return ret;
>   }
> +EXPORT_SYMBOL(rpmsg_register_device_override);
> +
> +int rpmsg_register_device(struct rpmsg_device *rpdev)
> +{
> +	return rpmsg_register_device_override(rpdev, NULL);
> +}
>   EXPORT_SYMBOL(rpmsg_register_device);
>   
>   /*
> diff --git a/drivers/rpmsg/rpmsg_internal.h b/drivers/rpmsg/rpmsg_internal.h
> index 3e81642238d2..a22cd4abe7d1 100644
> --- a/drivers/rpmsg/rpmsg_internal.h
> +++ b/drivers/rpmsg/rpmsg_internal.h
> @@ -94,19 +94,7 @@ int rpmsg_release_channel(struct rpmsg_device *rpdev,
>    */
>   static inline int rpmsg_ctrldev_register_device(struct rpmsg_device *rpdev)
>   {
> -	int ret;
> -
> -	strcpy(rpdev->id.name, "rpmsg_ctrl");
> -	ret = driver_set_override(&rpdev->dev, &rpdev->driver_override,
> -				  rpdev->id.name, strlen(rpdev->id.name));
> -	if (ret)
> -		return ret;
> -
> -	ret = rpmsg_register_device(rpdev);
> -	if (ret)
> -		kfree(rpdev->driver_override);
> -
> -	return ret;
> +	return rpmsg_register_device_override(rpdev, "rpmsg_ctrl");
>   }
>   
>   #endif
> diff --git a/drivers/rpmsg/rpmsg_ns.c b/drivers/rpmsg/rpmsg_ns.c
> index 8eb8f328237e..c70ad03ff2e9 100644
> --- a/drivers/rpmsg/rpmsg_ns.c
> +++ b/drivers/rpmsg/rpmsg_ns.c
> @@ -20,22 +20,10 @@
>    */
>   int rpmsg_ns_register_device(struct rpmsg_device *rpdev)
>   {
> -	int ret;
> -
> -	strcpy(rpdev->id.name, "rpmsg_ns");
> -	ret = driver_set_override(&rpdev->dev, &rpdev->driver_override,
> -				  rpdev->id.name, strlen(rpdev->id.name));
> -	if (ret)
> -		return ret;
> -
>   	rpdev->src = RPMSG_NS_ADDR;
>   	rpdev->dst = RPMSG_NS_ADDR;
>   
> -	ret = rpmsg_register_device(rpdev);
> -	if (ret)
> -		kfree(rpdev->driver_override);
> -
> -	return ret;
> +	return rpmsg_register_device_override(rpdev, "rpmsg_ns");
>   }
>   EXPORT_SYMBOL(rpmsg_ns_register_device);
>   
> diff --git a/include/linux/rpmsg.h b/include/linux/rpmsg.h
> index 20c8cd1cde21..523c98b96cb4 100644
> --- a/include/linux/rpmsg.h
> +++ b/include/linux/rpmsg.h
> @@ -165,6 +165,8 @@ static inline __rpmsg64 cpu_to_rpmsg64(struct rpmsg_device *rpdev, u64 val)
>   
>   #if IS_ENABLED(CONFIG_RPMSG)
>   
> +int rpmsg_register_device_override(struct rpmsg_device *rpdev,
> +				   const char *driver_override);
>   int rpmsg_register_device(struct rpmsg_device *rpdev);
>   int rpmsg_unregister_device(struct device *parent,
>   			    struct rpmsg_channel_info *chinfo);
> @@ -192,6 +194,12 @@ ssize_t rpmsg_get_mtu(struct rpmsg_endpoint *ept);
>   
>   #else
>   
> +static inline int rpmsg_register_device_override(struct rpmsg_device *rpdev,
> +						 const char *driver_override)
> +{
> +	return -ENXIO;
> +}
> +
>   static inline int rpmsg_register_device(struct rpmsg_device *rpdev)
>   {
>   	return -ENXIO;

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland