On 2015-09-07 08:42, Vladimir Laskov wrote:
> I have a problem with Linux guest on BHyve host. Guest use io
> driver virtio-blk.
>
>
> **
> # lsb_release -a
> No LSB modules are available.
> Distributor ID: Debi
Neel Natu wrote:
> Hi Paul,
>
> On Fri, Apr 10, 2015 at 10:35 AM, Paul Vixie wrote:
>> ...
>>
>> can we make TSC-low the default?
>>
>
> The choice of using the TSC is not without issues:
>
> - As rstone@ points out the TSCs need to be synchronized across physical cpus.
ok, then the reason i'm
Hi Paul,
On Fri, Apr 10, 2015 at 10:35 AM, Paul Vixie wrote:
> here is a representative one-hour sample of ntp messages from
> /var/log/messages on a freebsd 10.1 bhyve guest:
>
>> Apr 10 12:00:46 family ntpd[634]: time reset -0.613057 s
>> Apr 10 12:17:02 family ntpd[634]: time reset -0.604933 s
Hi,
On Fri, Apr 10, 2015 at 3:47 PM, Matthew Grooms wrote:
> Perhaps applying this commit would have some positive effect ( HPET vs LAPIC
> )?
>
> https://docs.freebsd.org/cgi/getmsg.cgi?fetch=885358+0+archive/2015/svn-src-head/20150322.svn-src-head
>
No, it's an orthogonal change.
The change a
On Fri, Apr 10, 2015 at 4:36 PM, Ryan Stone wrote:
> Using the TSC as the default timecounter in a VM is dangerous. On some
> hardware, the TSC is not synchronized across all CPU cores. This means
> that if a VM migrates from one core to another, it could see the
> timecounter value go backward
Perhaps applying this commit would have some positive effect ( HPET vs
LAPIC )?
https://docs.freebsd.org/cgi/getmsg.cgi?fetch=885358+0+archive/2015/svn-src-head/20150322.svn-src-head
-Matthew
On 4/10/2015 12:51 PM, Stephen Stuart wrote:
I think the right thing is to set (in /boot/loader.conf)
Using the TSC as the default timecounter in a VM is dangerous. On some
hardware, the TSC is not synchronized across all CPU cores. This means
that if a VM migrates from one core to another, it could see the
timecounter value go backwards. Time jumping backwards can cause all kinds
of hilarity.