[Acked/cmt] [xenial][PATCH v2 0/8] [Hyper-V] Implement Hyper-V PTP Source

Joshua R. Poulson jrp at pun.org
Mon Jul 3 12:32:25 UTC 2017


Userspace has to pick a PTP source, but we've generally recommended
using NTP and had it on by default in cloud images for years.

Thanks, --jrp

On Mon, Jul 3, 2017 at 5:18 AM, Andy Whitcroft <apw at canonical.com> wrote:
> On Fri, Jun 30, 2017 at 04:21:26PM -0300, Marcelo Henrique Cerri wrote:
>> BugLink: http://bugs.launchpad.net/bugs/1676635
>>
>> Adding a few more patches that fix additional issues introduced by the
>> original patch ("hv_utils: implement Hyper-V PTP source").
>>
>> v1:
>>
>> The requested commit requires more than 50 additional commits in order
>> to be applied cleanly. To avoid that, I cherry-picked just the strict
>> necessary commits and adapted to the Xenial kernel the upstream change
>> that makes the Hyper-V clock souce available via hyperv_cs.
>>
>> Dexuan Cui (1):
>>   Drivers: hv: util: don't forget to init host_ts.lock
>>
>> K. Y. Srinivasan (2):
>>   Drivers: hv: util: Use hv_get_current_tick() to get current tick
>>   Drivers: hv: util: Fix a typo
>>
>> Marcelo Henrique Cerri (1):
>>   UBUNTU: SAUCE: hv: make clocksource available for PTP device
>>     supporting
>>
>> Vitaly Kuznetsov (4):
>>   hv_util: switch to using timespec64
>>   hv_utils: implement Hyper-V PTP source
>>   hv_utils: drop .getcrosststamp() support from PTP driver
>>   hv_utils: fix TimeSync work on pre-TimeSync-v4 hosts
>>
>>  arch/x86/kernel/cpu/mshyperv.c |  23 ------
>>  drivers/hv/hv.c                |  36 ++++++++-
>>  drivers/hv/hv_util.c           | 168 +++++++++++++++++++++++++++++------------
>>  drivers/hv/hyperv_vmbus.h      |   3 +
>>  4 files changed, 158 insertions(+), 72 deletions(-)
>
> Well thankfully they are all self-contains to HV specific code.  The
> stack seems fairly small and benigh considering.  I feel this is going
> to put a burden on userspace to start using NTP.  Right?  I assume
> someone is sorting that out in userspace.
>
> Anyhow.  Overall it looks fine.  Assuming you do somethign sensible eith
> Klebers query:
>
> Acked-by: Andy Whitcroft <apw at canonical.com>
>
> -apw
>
> --
> kernel-team mailing list
> kernel-team at lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/kernel-team




More information about the kernel-team mailing list