Re: [PATCH] [tty/serial] 8250_acorn: Add error handling in serial_card_probe

[Hugo Villeneuve]

On Wed, 29 Nov 2023 03:52:36 -0800
Haoran Liu <liuhaoran14@xxxxxxx> wrote:

Hi,
you should add a proper prefix to your patch, like:
"serial: 8250_acorn: Add..."

You can use "git log --oneline drivers/tty/serial" to have an idea on
what prefix to add.

> This patch adds error handling to the serial_card_probe
> function in drivers/tty/serial/8250/8250_acorn.c. The

You can drop the full path to the file (and also the file itself) in
your commit log message, as this information is available in the diff.

> serial8250_register_8250_port call within this function
> previously lacked proper handling for failure scenarios.
> 
> Signed-off-by: Haoran Liu <liuhaoran14@xxxxxxx>
> ---
>  drivers/tty/serial/8250/8250_acorn.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/drivers/tty/serial/8250/8250_acorn.c b/drivers/tty/serial/8250/8250_acorn.c
> index 758c4aa203ab..378ae6936028 100644
> --- a/drivers/tty/serial/8250/8250_acorn.c
> +++ b/drivers/tty/serial/8250/8250_acorn.c
> @@ -43,6 +43,7 @@ serial_card_probe(struct expansion_card *ec, const struct ecard_id *id)
>  	struct uart_8250_port uart;
>  	unsigned long bus_addr;
>  	unsigned int i;
> +	int ret;
>  
>  	info = kzalloc(sizeof(struct serial_card_info), GFP_KERNEL);
>  	if (!info)
> @@ -72,6 +73,14 @@ serial_card_probe(struct expansion_card *ec, const struct ecard_id *id)
>  		uart.port.mapbase = bus_addr + type->offset[i];
>  
>  		info->ports[i] = serial8250_register_8250_port(&uart);
> +		if (IS_ERR(info->ports[i])) {
> +			ret = PTR_ERR(info->ports[i]);
> +			while (i--)
> +				serial8250_unregister_port(info->ports[i]);
> +
> +			kfree(info);
> +			return ret;

I am just wondering if unregistering all ports in case one fails is the
correct course of action? Looking at other drivers in the same folder
(8250_exar, 8250_pericom, 8250_pci), they seem to abort registering
new ports in case of error, but do not unregister previously registered
ports?

For 8250_pci1xxxx, in case of failure the for/loop still continues...

For 8250_men_mcb however, the probe exit in case of error.

Hugo Villeneuve.


> +		}
>  	}
>  
>  	return 0;
> -- 
> 2.17.1
> 
>