Re: [PATCH] use per-cpu variables in cpufreq

[Keir Fraser <keir.xen@gmail.com>]

On 30/05/2011 06:47, "Juergen Gross" <juergen.gross@ts.fujitsu.com> wrote:

>> Specifically, my fear is that this data gets pushed into the hypervisor
>> once-only during dom0 boot (via XENPF_set_processor_pminfo). If it is freed
>> during processor offline, we lose it forever and have no power management
>> when/if a CPU is brought back online. Worse I suspect your patch as it is
>> will crash if some CPUs are offline during boot as you'll deference their
>> per_cpu area which doesn't actually exist unless a CPU is online. You can
>> test this for yourself by adding a maxcpus=1 boot parameter for Xen.
> 
> Indeed.
> 
> Just to understand this completely:
> So when is this info set up for hot-plugged cpus? And what happens if
> a cpu module is removed and later replaced by another module with
> more cores (or threads) than the original one?
> I think the processor pminfo should be set in this case during the hotplug
> handling.

Well, there is a difference between logical and physical cpu hotplug. Xen is
capable of bringing CPUs online and offline without them actually being
physically plugged/unplugged from the mainboard. Indeed our physical hotplug
support is relatively new and I would suspect not much used (and it supports
only physical insertion, not removal!).

Frankly there are a number of questions around CPUs that are physically
plugged in after boot:
 * How does per-CPU ACPI state like PM info get set up?
 * In a system where TSCs are otherwise all perfectly in sync, does the
firmware help us by setting up the new CPUs' TSCs likewise?

I don't actually know the answers to these questions. Maybe dom0 ACPI does
get triggered on physical insertion and knows how to set up PM stuff?

So, when I say CPU hotplug, or online/offline, I'm generally focussing on
logical hotplug only still at the moment. Physical hotplug is a black art
with question marks hanging over it for me. ;-)

 -- Keir

>> The folding of the Intel/AMD structures might still be interesting, and
>> probably belongs as a separate patch anyway.
>> 
>> Cc'ing Intel and AMD guys to confirm this.
> 
> Okay, I'm waiting for their opinion.
> 
> 
> Juergen