[PATCH 2/2] nvme: keyring: fix conditional compilation

[Arnd Bergmann]

On Wed, Oct 25, 2023, at 17:00, Hannes Reinecke wrote:
> On 10/25/23 14:16, Arnd Bergmann wrote:
>> On Wed, Oct 25, 2023, at 11:11, Hannes Reinecke wrote:
>>> On 10/25/23 10:20, Arnd Bergmann wrote:
>
> 'IS_REACHABLE' is the keyword here; if we use that in nvme-keyring.h
> things are resolved.
>
> Will be sending an updated patch.

Please don't use IS_REACHABLE(), it's almost never the right
solution, even if it avoids this particular build failure,
it tends to cause hard to debug problems when features that
are enabled as modules end up not working correctly.

I still think my original patch or a variation of it is
the best option since it keys the usage of the module
directly on whether the caller actually wants to use it.

Another option would be to use Kconfig or Makefile logic
to force it to be built-in if at least one of the two
callers is built-in and at least one of them tries to
use it.

     Arnd