Re: [PATCH v2 1/2] module: Add a new helper delete_module()

[Petr Mladek <pmladek@suse.com>]

On Sun 2024-04-07 11:57:29, Yafang Shao wrote:
> Introduce a new helper function, delete_module(), designed to delete kernel
> modules from locations outside of the `kernel/module` directory.
> 
> No functional change.
> 
> --- a/kernel/module/main.c
> +++ b/kernel/module/main.c
> @@ -695,12 +695,74 @@ EXPORT_SYMBOL(module_refcount);
>  /* This exists whether we can unload or not */
>  static void free_module(struct module *mod);
>  
> +static void __delete_module(struct module *mod)
> +{
> +	char buf[MODULE_FLAGS_BUF_SIZE];
> +
> +	WARN_ON_ONCE(mod->state != MODULE_STATE_GOING);
> +
> +	/* Final destruction now no one is using it. */
> +	if (mod->exit != NULL)
> +		mod->exit();
> +	blocking_notifier_call_chain(&module_notify_list,
> +				     MODULE_STATE_GOING, mod);
> +	klp_module_going(mod);
> +	ftrace_release_mod(mod);
> +
> +	async_synchronize_full();
> +
> +	/* Store the name and taints of the last unloaded module for diagnostic purposes */
> +	strscpy(last_unloaded_module.name, mod->name, sizeof(last_unloaded_module.name));
> +	strscpy(last_unloaded_module.taints, module_flags(mod, buf, false),
> +		sizeof(last_unloaded_module.taints));
> +
> +	free_module(mod);
> +	/* someone could wait for the module in add_unformed_module() */
> +	wake_up_all(&module_wq);
> +}
> +
> +int delete_module(struct module *mod)
> +{
> +	int ret;
> +
> +	mutex_lock(&module_mutex);
> +	if (!list_empty(&mod->source_list)) {
> +		/* Other modules depend on us: get rid of them first. */
> +		ret = -EWOULDBLOCK;
> +		goto out;
> +	}

This is cut&paste from SYSCALL_DEFINE2(delete_module...

> +
> +	/* Doing init or already dying? */
> +	if (mod->state != MODULE_STATE_LIVE) {
> +		ret = -EBUSY;
> +		goto out;
> +	}

Same here. You only removed the debug message. Why?

> +
> +	/* If it has an init func, it must have an exit func to unload */
> +	if (mod->init && !mod->exit) {
> +		ret = -EBUSY;
> +		goto out;
> +	}

Same code, just without the "forced" handling.

> +
> +	if (try_release_module_ref(mod) != 0) {
> +		ret = -EWOULDBLOCK;
> +		goto out;
> +	}

This is the same as try_stop_module() without the "forced" handling.

> +	mod->state = MODULE_STATE_GOING;
> +	mutex_unlock(&module_mutex);
> +	__delete_module(mod);
> +	return 0;

I am sure that we could better refactor the code to remove
the code duplication.

> +
> +out:
> +	mutex_unlock(&module_mutex);
> +	return ret;
> +}
> +
>  SYSCALL_DEFINE2(delete_module, const char __user *, name_user,
>  		unsigned int, flags)
>  {
>  	struct module *mod;
>  	char name[MODULE_NAME_LEN];
> -	char buf[MODULE_FLAGS_BUF_SIZE];
>  	int ret, forced = 0;
>  
>  	if (!capable(CAP_SYS_MODULE) || modules_disabled)

Otherwise, it looks good to me.

Best Regards,
Petr