Re: [PATCH] module.h: allow #define strings to work with MODULE_IMPORT_NS

[Matthias Maennich]

On Sat, Jan 08, 2022 at 03:06:57PM +0100, Greg Kroah-Hartman wrote:
> The MODULE_IMPORT_NS() macro does not allow defined strings to work
> properly with it, so add a layer of indirection to allow this to happen.
>
> Cc: Luis Chamberlain <mcgrof@xxxxxxxxxx>
> Cc: Jessica Yu <jeyu@xxxxxxxxxx>
> Cc: Matthias Maennich <maennich@xxxxxxxxxx>
> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

Reviewed-by: Matthias Maennich <maennich@xxxxxxxxxx>

Cheers,
Matthias

> ---
> Note, there is no in-kernel user of this, I wrote this to help deal with
> some module namespace issues that Android kernels were having where we
> wanted to put the namespace name in a build #define to make things
> easier over time when combined with the change at:
> 	https://lore.kernel.org/all/20220108140415.3360088-1-gregkh@xxxxxxxxxxxxxxxxxxx/
>
> I can also see this be used in ways to make namespace names up "on the
> fly at build time" when combined with the above-referenced change to
> create a different namespace for every time the kernel is built to do a
> poor-man's symbol mangling to make life harder for out-of-tree modules.
> Not that I would ever suggest such a thing :)
>
> include/linux/module.h | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/include/linux/module.h b/include/linux/module.h
> index c9f1200b2312..f4338235ed2c 100644
> --- a/include/linux/module.h
> +++ b/include/linux/module.h
> @@ -290,7 +290,8 @@ extern typeof(name) __mod_##type##__##name##_device_table		\
>  * files require multiple MODULE_FIRMWARE() specifiers */
> #define MODULE_FIRMWARE(_firmware) MODULE_INFO(firmware, _firmware)
>
> -#define MODULE_IMPORT_NS(ns) MODULE_INFO(import_ns, #ns)
> +#define _MODULE_IMPORT_NS(ns)	MODULE_INFO(import_ns, #ns)
> +#define MODULE_IMPORT_NS(ns)	_MODULE_IMPORT_NS(ns)
>
> struct notifier_block;
>
> --
> 2.34.1