Re: Early microcode signing in secure boot environment - Was: x86, microcode: Use common get_ramdisk_image()

From: Thomas Renninger
Date: Thu Apr 11 2013 - 03:31:40 EST


On Wednesday, April 10, 2013 05:47:25 PM Yu, Fenghua wrote:
> > -----Original Message-----
> > From: Thomas Renninger [mailto:trenn@xxxxxxx]
> > Sent: Wednesday, April 10, 2013 12:41 AM
> > Hello,
> >
> > On Wednesday, April 10, 2013 01:34:33 PM Tang Chen wrote:
> > > On 04/05/2013 07:46 AM, Yinghai Lu wrote:
> > > > Use common get_ramdisk_image() to get ramdisk start phys address.
> > > >
> > > > We need this to get correct ramdisk adress for 64bit bzImage that
> > > > initrd can be loaded above 4G by kexec-tools.disk_size;
> >
> > don't know whether this question came up when this feature got
> > submitted (if yes a pointer would be appreciated).
> >
> > Is there a concept how signed microcode can get verified when applied
> > early,
> > like it is done via firmware loader?
> >
> > If not, early microcode loading is not really usable in secure boot
> > environment, right?
>
> The microcode is cryptographically authenticated by the CPU itself, so there
> is no security issue related to early microcode loading.

So Intel HW is allowed to authenticate its firmware itself, bypassing the UEFI
certificates...
Does this apply for other vendors as well?
Does this apply to secure boot specification?

Is this "cryptographically authenticated by the CPU itself" thing documented
somewhere so that security people can double check that it is really
secure?

Just some questions to discuss and think about...

Thomas
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/