On Wed, Sep 05, 2018 at 08:59:24PM +0800, Pu Wen wrote:
> If CONFIG_X86_MCE_AMD=n, mce_hygon_feature_init will call the other
> one mce_amd_feature_init which is a null function and located in the
> else branch of "#ifdef CONFIG_X86_MCE_AMD". The compilation is OK and
> the kernel will run without MCE support.
So my question was rather ironic but I guess irony can't travel through
mail. So let me paraphrase: is that a use case you wanna support?
Because I'd advise very strongly against !MCE kernels.
--
Regards/Gruss,
Boris.
Good mailing practices for 400: avoid top-posting and trim the reply.
