On Thu, Mar 05, 2015 at 02:24:17PM +0300, Dan Carpenter wrote:
> Static checkers complain about the missing call to misc_deregister() if
> bind_virq_for_mce() fails.
>
> Also I reversed the tests so that we do error handling instead of
> success handling. That way we just have a series of function calls
> instead of the more complicated nested if statements in the original
> code. Let's preserve the error codes as well.
>
> Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>
David, you want me to stash in devel for 4.1?
>
> diff --git a/drivers/xen/mcelog.c b/drivers/xen/mcelog.c
> index 6ab6a79..a493c7315 100644
> --- a/drivers/xen/mcelog.c
> +++ b/drivers/xen/mcelog.c
> @@ -393,14 +393,25 @@ static int bind_virq_for_mce(void)
>
> static int __init xen_late_init_mcelog(void)
> {
> + int ret;
> +
> /* Only DOM0 is responsible for MCE logging */
> - if (xen_initial_domain()) {
> - /* register character device /dev/mcelog for xen mcelog */
> - if (misc_register(&xen_mce_chrdev_device))
> - return -ENODEV;
> - return bind_virq_for_mce();
> - }
> + if (!xen_initial_domain())
> + return -ENODEV;
> +
> + /* register character device /dev/mcelog for xen mcelog */
> + ret = misc_register(&xen_mce_chrdev_device);
> + if (ret)
> + return ret;
> +
> + ret = bind_virq_for_mce();
> + if (ret)
> + goto deregister;
>
> - return -ENODEV;
> + return 0;
> +
> +deregister:
> + misc_deregister(&xen_mce_chrdev_device);
> + return ret;
> }
> device_initcall(xen_late_init_mcelog);
--
To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
