On 06/28/14 22:07, Stephan Mueller wrote:
> Am Sonntag, 29. Juni 2014, 12:20:15 schrieb Stephen Rothwell:
>
> Hi Stephen,
>
>> Hi Stephan,
>>
>> On Sat, 28 Jun 2014 22:00:07 +0200 Stephan Mueller <smueller@xxxxxxxxxx>
> wrote:
>>> diff --git a/crypto/drbg.c b/crypto/drbg.c
>>> index 6679a26..03a230e 100644
>>> --- a/crypto/drbg.c
>>> +++ b/crypto/drbg.c
>>> @@ -102,8 +102,13 @@
>>>
>>> #if !defined(CONFIG_CRYPTO_DRBG_HASH) && \
>>>
>>> !defined(CONFIG_CRYPTO_DRBG_HMAC) && \
>>> !defined(CONFIG_CRYPTO_DRBG_CTR)
>>>
>>> -#warning "The DRBG code is useless without compiling at least one DRBG
>>> type" -#endif
>>> +#define CRYPTO_DRBG_NONE_STRING "none "
>>> +static int __init drbg_init(void)
>>> +{
>>> + pr_warn("DRBG: no DRBG core was compiled!\n");
>>> + return -EFAULT;
>>> +}
>>> +#else
>>
>> Wouldn't this be better handled by Kconfig so that we don't even try to
>> build this unless one of the required core modules is chosen?
>
> I tried that, but it seems that my Kconfig Foo is not too well: adding the
> DRBG cores to the depends line of CRYPTO_DRBG as indicated in the following, I
> have a circular dependency. With that circular dependency, the DRBG entries do
> not show up in make menuconfig.
>
> menuconfig CRYTPO_DRBG
> tristate "NIST SP800-90A DRBG"
> depends on CRYPTO && (CRYPTO_DRBG_HMAC || CRYPTO_DRBG_CTR ||
> CRYPTO_DRBG_HASH)
> ...
>
> if CRYTPO_DRBG
>
> config CRYPTO_DRBG_HMAC
> bool "Enable HMAC DRBG"
> default y
> depends on CRYTPO_DRBG
>
> Do you have a working solution in mind? The goal is that once CRYPTO_DRBG is
> selected, at least one of the DRBG cores must be selected.
That sounds like a 'choice' Kconfig could be used.
Have you looked at that possibility?
See Documentation/kbuild/kconfig-language.txt and search for 'choice'.
--
~Randy
--
To unsubscribe from this list: send the line "unsubscribe linux-next" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
