Re: [Xen-devel] [PATCH v4 2/6] microcode: save all microcodes which pass sanity check

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 30.11.18 at 08:55, <chao.gao@xxxxxxxxx> wrote:
> On Thu, Nov 29, 2018 at 10:22:10AM +0100, Roger Pau Monné wrote:
>>On Thu, Nov 29, 2018 at 10:40:32AM +0800, Chao Gao wrote:
>>> On Wed, Nov 28, 2018 at 01:00:14PM +0100, Roger Pau Monné wrote:
>>> >On Wed, Nov 28, 2018 at 01:34:12PM +0800, Chao Gao wrote:
>>> >> ... and search caches to find a suitable one when loading.
>>> >
>>> >Why do you need to save all of them? You are only going to load a
>>> >single microcode, so I don't understand the need to cache them all.
>>
>>I think the above question needs an answer.
> 
> Out of consideraton for a mixed-family system. Anyway, Since Jan commented
> that we gave up support of a mixed-family system, we only need to save
> a single microcode for offlined or hot-plugged cpus.

Well, there might be slightly more needed than just a single instance.
This depends on whether same family/model CPUs with different
stepping and/or different "pf" value can be mixed (and would have
identical feature flags in their CPUID output).

I may have oversimplified the current state of things: Hotplugging
a CPU with more capabilities than the boot CPU is generally fine
afaict. Both you (Intel) and AMD place restrictions on permitted
mixes iirc, so I don't think we need to support anything beyond
such restrictions. Being able to run on permitted mixes which are
not in conflict with our general assumption of there not being any
CPU in the system with less capabilities than the boot CPU would
seem desirable.

Andrew, do you have any view or opinion here?

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel