[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] MCE/EDAC Status/Updating?

>>> On 18.02.19 at 09:42, <ehem+xen@xxxxxxx> wrote:
> On Mon, Feb 18, 2019 at 01:12:16AM -0700, Jan Beulich wrote:
>> >>> On 15.02.19 at 19:20, <ehem+xen@xxxxxxx> wrote:
>> > On Fri, Feb 15, 2019 at 03:58:49AM -0700, Jan Beulich wrote:
>> >> Well, Fam10 is mentioned explicitly, but as per the use of e.g.
>> >> mcheck_amd_famXX newer ones are supported by this code
>> >> as well.
>> > 
>> > In that case sometime between Xen 4.1 and Xen 4.4, the AMD MCE/EDAC code
>> > was completely broken and hasn't been fixed.
>> I can't say I'm surprised, but details of the breakage would still
>> be appreciated.
> Originally noticed with Debian: https://bugs.debian.org/810964 
> Original observer noticed that half the memory controllers were missing
> from Linux's Domain-0 dmesg with Xen 4.4.  EDAC capability flags are
> missing with Xen 4.4.

And I had been commenting in this bug. I don't recall technical data
ever having emerged on the list here as to what is really going on,
and what the root of this perceived regression is.

> I'd been working with a processor Linux was reporting as
> "cpu family : 16" (ah yes "10h", that funky olde way of refering to
> things) and noticing Linux's EDAC support failing on kernel start.  In
> which case the EDAC support on AMD processors was completely broken
> between 4.1 and 4.4 (hadn't realized that processor was just old enough
> to be interesting).

While there's a relation, I think we need to keep #MC handling
and EDAC separate here: The latter lives entirely in Dom0. And
as said in the Debian bug, at least back at the time there was no
reason to believe the driver would work on Xen other than by


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.