Re: [PATCH] soc: aspeed: Add NULL pointer check in aspeed_lpc_enable_snoop()

[Andrew Lunn]

On Wed, Feb 12, 2025 at 07:50:49PM -0600, Chenyuan Yang wrote:
> Hi Andrew,
> 
> I've drafted the following patch to address the resource cleanup issue:

Please just follow the normal procedure of submitting a patch.

	Andrew

> 
> ```
>  drivers/soc/aspeed/aspeed-lpc-snoop.c | 17 +++++++++++++++--
>  1 file changed, 15 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/soc/aspeed/aspeed-lpc-snoop.c
> b/drivers/soc/aspeed/aspeed-lpc-snoop.c
> index 9ab5ba9cf1d6..4988144aba88 100644
> --- a/drivers/soc/aspeed/aspeed-lpc-snoop.c
> +++ b/drivers/soc/aspeed/aspeed-lpc-snoop.c
> @@ -200,11 +200,15 @@ static int aspeed_lpc_enable_snoop(struct
> aspeed_lpc_snoop *lpc_snoop,
>   lpc_snoop->chan[channel].miscdev.minor = MISC_DYNAMIC_MINOR;
>   lpc_snoop->chan[channel].miscdev.name =
>   devm_kasprintf(dev, GFP_KERNEL, "%s%d", DEVICE_NAME, channel);
> + if (!lpc_snoop->chan[channel].miscdev.name) {
> + rc = -ENOMEM;
> + goto free_fifo;
> + }

You were asked to first add cleanup, then fix this possible NULL
pointer dereference.

> I have a couple of questions regarding the cleanup order:
> 
> 1. Do we need to call misc_deregister() in this case, considering that
> the registration happens before return -EINVAL?
> 2. If misc_deregister() is required, is there a specific order we
> should follow when calling misc_deregister() and
> kfree(lpc_snoop->chan[channel].miscdev.name);?

As a general rule, cleanup is the opposite order to setup.

Also, you want to do some research about that devm_ means.

	Andrew