Re: [PATCH v1] bcache: Check for NULL return of kzalloc()

From: Coly Li <colyli@suse.de>
To: QintaoShen <unSimple1993@163.com>
Cc: linux-bcache@vger.kernel.org, kent.overstreet@gmail.com,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v1] bcache: Check for NULL return of kzalloc()
Date: Fri, 8 Apr 2022 00:59:34 +0800	[thread overview]
Message-ID: <c934c457-fe8f-7937-f348-eccf46375878@suse.de> (raw)
In-Reply-To: <1648114074-10045-1-git-send-email-unSimple1993@163.com>

On 3/24/22 5:27 PM, QintaoShen wrote:
> kzalloc() is a memory allocation function which may return a NULL pointer.
> Therefore, it is better to check the return value of kzalloc() to avoid potential
> NULL-pointer dereference.
>
> Signed-off-by: QintaoShen <unSimple1993@163.com>
> ---
>   drivers/md/bcache/request.c | 3 +++
>   1 file changed, 3 insertions(+)
>
> diff --git a/drivers/md/bcache/request.c b/drivers/md/bcache/request.c
> index fdd0194..232ffe3 100644
> --- a/drivers/md/bcache/request.c
> +++ b/drivers/md/bcache/request.c
> @@ -1105,6 +1105,9 @@ static void detached_dev_do_request(struct bcache_device *d, struct bio *bio,
>   	 * which would call closure_get(&dc->disk.cl)
>   	 */
>   	ddip = kzalloc(sizeof(struct detached_dev_io_private), GFP_NOIO);
> +	if (!ddpp)
> +		return ;

This is wrong. If you test your patch, you may experience a system hang 
I guess.

Coly Li

> +
>   	ddip->d = d;
>   	/* Count on the bcache device */
>   	ddip->orig_bdev = orig_bdev;