On 25.11.2019 11:15, Jan Beulich wrote:
On 23.11.2019 00:10, Andreas Kinzler wrote:
BTW: Xeon E-2136 @ C242 has 8086:3eca as ID. One needs to check with
Intel which combinations are really affected.
Are you saying you observed the same issue on such a (server processor)
system as well? Neither its datasheet nor its specification update
The whole thread starting with
https://lists.xenproject.org/archives/html/xen-devel/2019-10/msg00966.html
was about Xeon E-2136.
Setting a limit to PC7 greatly reduced the drift
(https://lists.xenproject.org/archives/html/xen-devel/2019-11/msg01044.html)
(which I specifically downloaded and looked through just because of your
remark) have any mention of a similar issue. I also take it that the
code comment inherited from Linux says "SoCs" for a reason.
Even the kernel mailing list postings lack official confirmation from
Intel. That is why I said: someone (with internal Intel knowledge) needs
to confirm which combinations are affected.
Regards Andreas
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel