Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Date:   Fri, 11 Mar 2022 14:54:46 -0600
From:   Bjorn Andersson <bjorn.andersson@linaro.org>
To:     Deepak Kumar Singh <quic_deesin@quicinc.com>
Cc:     swboyd@chromium.org, quic_clew@quicinc.com,
        mathieu.poirier@linaro.org, linux-kernel@vger.kernel.org,
        linux-arm-msm@vger.kernel.org, linux-remoteproc@vger.kernel.org,
        Ohad Ben-Cohen <ohad@wizery.com>
Subject: Re: [PATCH V1 3/3] rpmsg: glink: Add lock for ctrl device
Message-ID: <Yiu3Fl/Diw5iqh24@builder.lan>
References: <1643223886-28170-1-git-send-email-quic_deesin@quicinc.com>
 <1643223886-28170-4-git-send-email-quic_deesin@quicinc.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1643223886-28170-4-git-send-email-quic_deesin@quicinc.com>
Precedence: bulk

On Wed 26 Jan 13:04 CST 2022, Deepak Kumar Singh wrote:

> Race between rpmsg_eptdev_create and rpmsg_chrdev_remove
> can sometime casue crash while accessing rpdev while new
> endpoint is being created. Using lock ensure no new eptdev
> is created after rpmsg_chrdev_remove has been completed.

This patch lacks a Signed-off-by.

Isn't this solving the same problem as the previous patch? Would be nice
with some more specifics on the race that you're seeing.

Thanks,
Bjorn

> ---
>  drivers/rpmsg/rpmsg_char.c | 11 ++++++++++-
>  1 file changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/rpmsg/rpmsg_char.c b/drivers/rpmsg/rpmsg_char.c
> index 2108ef8..3e5b85d 100644
> --- a/drivers/rpmsg/rpmsg_char.c
> +++ b/drivers/rpmsg/rpmsg_char.c
> @@ -27,6 +27,7 @@
>  
>  static dev_t rpmsg_major;
>  static struct class *rpmsg_class;
> +struct mutex ctrl_lock;
>  
>  static DEFINE_IDA(rpmsg_ctrl_ida);
>  static DEFINE_IDA(rpmsg_ept_ida);
> @@ -396,9 +397,12 @@ static int rpmsg_eptdev_create(struct rpmsg_ctrldev *ctrldev,
>  	struct device *dev;
>  	int ret;
>  
> +	mutex_lock(&ctrl_lock);
>  	eptdev = kzalloc(sizeof(*eptdev), GFP_KERNEL);
> -	if (!eptdev)
> +	if (!eptdev) {
> +		mutex_unlock(&ctrl_lock);
>  		return -ENOMEM;
> +	}
>  
>  	dev = &eptdev->dev;
>  	eptdev->rpdev = rpdev;
> @@ -443,6 +447,7 @@ static int rpmsg_eptdev_create(struct rpmsg_ctrldev *ctrldev,
>  		put_device(dev);
>  	}
>  
> +	mutex_unlock(&ctrl_lock);
>  	return ret;
>  
>  free_ept_ida:
> @@ -453,6 +458,7 @@ static int rpmsg_eptdev_create(struct rpmsg_ctrldev *ctrldev,
>  	put_device(dev);
>  	kfree(eptdev);
>  
> +	mutex_unlock(&ctrl_lock);
>  	return ret;
>  }
>  
> @@ -525,6 +531,7 @@ static int rpmsg_chrdev_probe(struct rpmsg_device *rpdev)
>  	if (!ctrldev)
>  		return -ENOMEM;
>  
> +	mutex_init(&ctrl_lock);
>  	ctrldev->rpdev = rpdev;
>  
>  	dev = &ctrldev->dev;
> @@ -581,12 +588,14 @@ static void rpmsg_chrdev_remove(struct rpmsg_device *rpdev)
>  	int ret;
>  
>  	/* Destroy all endpoints */
> +	mutex_lock(&ctrl_lock);
>  	ret = device_for_each_child(&ctrldev->dev, NULL, rpmsg_eptdev_destroy);
>  	if (ret)
>  		dev_warn(&rpdev->dev, "failed to nuke endpoints: %d\n", ret);
>  
>  	device_del(&ctrldev->dev);
>  	put_device(&ctrldev->dev);
> +	mutex_unlock(&ctrl_lock);
>  }
>  
>  static struct rpmsg_driver rpmsg_chrdev_driver = {
> -- 
> 2.7.4
>