Re: [PATCH 0/2] maintaining of kvm_clock stability on guest vCPU hotplug

Re: [PATCH 0/2] maintaining of kvm_clock stability on guest vCPU hotplug

[Denis Plotnikov]

ping

On 23.03.2017 17:28, Denis Plotnikov wrote:
> It was found, that after adding a new vCPU to a running VM (CPU hotplug)
> kvm_clock lost the stability property in case of using it. This happened
> because KVM didn't treated cpu hotplug as a special case of TSC
> synchronization.
> This patch series adds cpu hotplug to the TSC synchronization cases.
>
>
> Denis Plotnikov (2):
>   KVM: x86: remaster kvm_write_tsc code
>   KVM: x86: fix maintaining of kvm_clock stability on guest CPU hotplug
>
>  arch/x86/kvm/x86.c | 59 ++++++++++++++++++------------------------------------
>  1 file changed, 20 insertions(+), 39 deletions(-)
>

-- 
Best,
Denis

Re: [PATCH 0/2] maintaining of kvm_clock stability on guest vCPU hotplug

[Radim Krčmář]

2017-04-05 11:56+0300, Denis Plotnikov:
> ping

Please re-send the series, I don't have it in mailbox and web-archives
don't show it either,

thanks.

> On 23.03.2017 17:28, Denis Plotnikov wrote:
>> It was found, that after adding a new vCPU to a running VM (CPU hotplug)
>> kvm_clock lost the stability property in case of using it. This happened
>> because KVM didn't treated cpu hotplug as a special case of TSC
>> synchronization.
>> This patch series adds cpu hotplug to the TSC synchronization cases.

Right, it looks like an incarnation of the deep problem where the kvm
clock with masterclock diverges from host (different frequency), so any
case where the masterclock is synchronized results in time shifts in the
guest.

>> 
>> Denis Plotnikov (2):
>>   KVM: x86: remaster kvm_write_tsc code
>>   KVM: x86: fix maintaining of kvm_clock stability on guest CPU hotplug
>> 
>>  arch/x86/kvm/x86.c | 59 ++++++++++++++++++------------------------------------
>>  1 file changed, 20 insertions(+), 39 deletions(-)
>> 
> 
> -- 
> Best,
> Denis

Re: [PATCH 0/2] maintaining of kvm_clock stability on guest vCPU hotplug

[Denis Plotnikov]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="y", Size: 1124 bytes --]

the series sending repeated

On 05.04.2017 17:23, Radim Krčmář wrote:
> 2017-04-05 11:56+0300, Denis Plotnikov:
>> ping
> Please re-send the series, I don't have it in mailbox and web-archives
> don't show it either,
> 
> thanks.
> 
>> On 23.03.2017 17:28, Denis Plotnikov wrote:
>>> It was found, that after adding a new vCPU to a running VM (CPU hotplug)
>>> kvm_clock lost the stability property in case of using it. This happened
>>> because KVM didn't treated cpu hotplug as a special case of TSC
>>> synchronization.
>>> This patch series adds cpu hotplug to the TSC synchronization cases.
> Right, it looks like an incarnation of the deep problem where the kvm
> clock with masterclock diverges from host (different frequency), so any
> case where the masterclock is synchronized results in time shifts in the
> guest.
> 
>>> Denis Plotnikov (2):
>>>   KVM: x86: remaster kvm_write_tsc code
>>>   KVM: x86: fix maintaining of kvm_clock stability on guest CPU hotplug
>>>
>>>  arch/x86/kvm/x86.c | 59 ++++++++++++++++++------------------------------------
>>>  1 file changed, 20 insertions(+), 39 deletions(-)
>>>