On 01/21/2014 11:14 AM, Borislav Petkov wrote:
> On Tue, Jan 21, 2014 at 10:39:21AM -0500, Boris Ostrovsky wrote:
>> I was thinking this could have been introduced by last night's merge
>> of your tree.
> Ok, I need more info:
>
> * kernel .config

Below, but probably not particularly important because...

>
> * are you building an initrd with the microcode? If so, how exactly?
> Which microcode blob?

         cat $(AMD_UCODE_PATH)/* > 
ucode_initrd/kernel/x86/microcode/AuthenticAMD.bin
         (cd ucode_initrd;find . | cpio -o -H newc > 
../$(DISTDIR)/common/ucode_initrd.cpio)
         ...
         cat $(DISTDIR)/common/ucode_initrd.cpio \
             $(DISTDIR)/common/_initramfs.cpio.gz > \
             $(DISTDIR)/common/initramfs.cpio.gz

and as I just discovered, with a little twist: apparently AMD_UCODE_PATH 
points to Intel microcode.

So we clearly screwed up on our end but I'd think that we shouldn't 
crash when this happens. We haven't until today (and evidently we've had 
this problem for a long time).

>
> * what machine, family?

All of our AMD boxes. E.g.:

CPU0: AMD E-350 Processor (fam: 14, model: 01, stepping: 00)
CPU0: Quad-Core AMD Opteron(tm) Processor 1352 (fam: 10, model: 02, 
stepping: 03)
CPU0: AMD Engineering Sample (fam: 10, model: 04, stepping: 01)
CPU0: AMD Athlon(tm) II X2 245 Processor (fam: 10, model: 06, stepping: 02)
CPU0: AMD A8-3850 APU with Radeon(tm) HD Graphics (fam: 12, model: 01, 
stepping: 00)

>
> * anything else I'd need to reproduce here?
>
> Btw, why the h*ll are you still using a 32-bit kernel? :)

It's part of Xen testing and 32-bit dom0 is still being used. And 
because we build it we test bare-metal as well.

-boris