[RFC] Handling of number of cores the guest sees

[RFC] Handling of number of cores the guest sees

[Andre Przywara]

Hi,

while experimenting with guest NUMA configurations I realized that Xen 
injects the host's core number into each guest.
I believe this behavior is wrong, the number of cores should somehow be 
dependent from the number of VCPUs.
Currently a CPUID decoding tool of mine gives me the following output 
for a 4 VCPU guest:
------------
HTT: 1, CmpLegacy: 1, LogicalProcessorCount: 24
AMD decoding:
         NC: 23, ApicIdCoreIdSize: 16
             24 cores   (legacy method)
          24/16 cores (extended method)
------------
(This is on a 12-core host CPU).
Applying my previous patch reduces the 24 to 12, but that still does not 
match the 4 VCPUs seen.
For proper NUMA functionality we need more sane values here, it seems 
that at least Linux does not care about the strange numbers as long as 
NUMA is not used. When a SRAT table is found, the guest kernel panics in 
the scheduler's rebalancer with those bogus numbers.

How shall we solve this issue? I see several ways:

1. Always inject one core per processor. SMP guests are then n-way, the 
CPUID setup is trivial and works well. But we may run into licensing 
issues, as some software (MS Windows comes to mind) is limited by the 
number of processors, but not by the number of cores.

2. Inject exactly the same number of cores as there are VCPUs. This 
could lead to potentially strange core numbers, but software should cope 
with this (as there are 3-core, 6-core and 12-cores processors).
This would lead to problems with a NUMA setup, though.

3. Let the user specify the number of cores in the config file. Needs 
user interaction and can lead to problems if it somehow conflicts the 
number of VCPUs. But would be nice to have as an additional tuning 
parameter. I could implement this.

4. Implement solution 2), but tune the behavior if guest NUMA is 
enabled. We could make sure that the number of cores is not bigger than 
the number of VCPUs on one NUMA node.

What approach shall I use? Are there other concerns regarding the 
CPUID's readout of the nubmer of cores?

Regards,
Andre.

-- 
Andre Przywara
AMD-Operating System Research Center (OSRC), Dresden, Germany
Tel: +49 351 448-3567-12

Re: [RFC] Handling of number of cores the guest sees

[Keir Fraser]

On 25/08/2010 13:06, "Andre Przywara" <andre.przywara@amd.com> wrote:

> 1. Always inject one core per processor. SMP guests are then n-way, the
> CPUID setup is trivial and works well. But we may run into licensing
> issues, as some software (MS Windows comes to mind) is limited by the
> number of processors, but not by the number of cores.

Yeah that was why the hack in cpuid logic was put in. It ought to be smarter
and correcter than it is though. Personally I'd be happy for it to go, at
least in the case that you run with NUMA passthru.

 -- Keir

RE: [RFC] Handling of number of cores the guest sees

[Huang2, Wei]

(1) is the issue we try to resolve. So we should avoid it as much as possible.

I would prefer an approach to combine (2) and (3) together. Users can specify vcpus and cores_per_cpu in guest configure file. Xen will automatically parse and setup it up. For instance, vcpus=7 and cores_per_cpu=4 will present guest VM with 2 physical CPUs: one with quad cores and one with triple cores. Most modern OS can deal with odd core numbers.

-Wei
-----Original Message-----
From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-bounces@lists.xensource.com] On Behalf Of Andre Przywara
Sent: Wednesday, August 25, 2010 7:07 AM
To: Keir Fraser; Nitin Kamble
Cc: xen-devel
Subject: [Xen-devel] [RFC] Handling of number of cores the guest sees

Hi,

while experimenting with guest NUMA configurations I realized that Xen 
injects the host's core number into each guest.
I believe this behavior is wrong, the number of cores should somehow be 
dependent from the number of VCPUs.
Currently a CPUID decoding tool of mine gives me the following output 
for a 4 VCPU guest:
------------
HTT: 1, CmpLegacy: 1, LogicalProcessorCount: 24
AMD decoding:
         NC: 23, ApicIdCoreIdSize: 16
             24 cores   (legacy method)
          24/16 cores (extended method)
------------
(This is on a 12-core host CPU).
Applying my previous patch reduces the 24 to 12, but that still does not 
match the 4 VCPUs seen.
For proper NUMA functionality we need more sane values here, it seems 
that at least Linux does not care about the strange numbers as long as 
NUMA is not used. When a SRAT table is found, the guest kernel panics in 
the scheduler's rebalancer with those bogus numbers.

How shall we solve this issue? I see several ways:

1. Always inject one core per processor. SMP guests are then n-way, the 
CPUID setup is trivial and works well. But we may run into licensing 
issues, as some software (MS Windows comes to mind) is limited by the 
number of processors, but not by the number of cores.

2. Inject exactly the same number of cores as there are VCPUs. This 
could lead to potentially strange core numbers, but software should cope 
with this (as there are 3-core, 6-core and 12-cores processors).
This would lead to problems with a NUMA setup, though.

3. Let the user specify the number of cores in the config file. Needs 
user interaction and can lead to problems if it somehow conflicts the 
number of VCPUs. But would be nice to have as an additional tuning 
parameter. I could implement this.

4. Implement solution 2), but tune the behavior if guest NUMA is 
enabled. We could make sure that the number of cores is not bigger than 
the number of VCPUs on one NUMA node.

What approach shall I use? Are there other concerns regarding the 
CPUID's readout of the nubmer of cores?

Regards,
Andre.

-- 
Andre Przywara
AMD-Operating System Research Center (OSRC), Dresden, Germany
Tel: +49 351 448-3567-12


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel