Linux Virtualization - Sep 2019 - [PATCH 0/2] Revert and rework on the metadata accelreation

On 2019/9/5 ??9:59, Jason Gunthorpe wrote:
> On Thu, Sep 05, 2019 at 08:27:34PM +0800, Jason Wang wrote:
>> Hi:
>>
>> Per request from Michael and Jason, the metadata accelreation is
>> reverted in this version and rework in next version.
>>
>> Please review.
>>
>> Thanks
>>
>> Jason Wang (2):
>>    Revert "vhost: access vq metadata through kernel virtual
address"
>>    vhost: re-introducing metadata acceleration through kernel virtual
>>      address
> There are a bunch of patches in the queue already that will help
> vhost, and I a working on one for next cycle that will help alot more
> too.

I will check those patches, but if you can give me some pointers or 
keywords it would be much appreciated.

>
> I think you should apply the revert this cycle and rebase the other
> patch for next..
>
> Jason

Yes, the plan is to revert in this release cycle.

Thanks

From: Jason Wang <jasowang at redhat.com>
Date: Fri, 6 Sep 2019 18:02:35 +0800
> On 2019/9/5 ??9:59, Jason Gunthorpe wrote:
>> I think you should apply the revert this cycle and rebase the other
>> patch for next..
>>
>> Jason
> 
> Yes, the plan is to revert in this release cycle.
Then you should reset patch #1 all by itself targetting 'net'.

On Fri, Sep 06, 2019 at 06:02:35PM +0800, Jason Wang
wrote:
> 
> On 2019/9/5 ??9:59, Jason Gunthorpe wrote:
> > On Thu, Sep 05, 2019 at 08:27:34PM +0800, Jason Wang wrote:
> > > Hi:
> > > 
> > > Per request from Michael and Jason, the metadata accelreation is
> > > reverted in this version and rework in next version.
> > > 
> > > Please review.
> > > 
> > > Thanks
> > > 
> > > Jason Wang (2):
> > >    Revert "vhost: access vq metadata through kernel virtual
address"
> > >    vhost: re-introducing metadata acceleration through kernel
virtual
> > >      address
> > There are a bunch of patches in the queue already that will help
> > vhost, and I a working on one for next cycle that will help alot more
> > too.
> 
> 
> I will check those patches, but if you can give me some pointers or
keywords
> it would be much appreciated.
You can look here:

https://github.com/jgunthorpe/linux/commits/mmu_notifier

The first parts, the get/put are in the hmm tree, and the last part,
the interval tree in the last commit is still a WIP, but it would
remove alot of that code from vhost as well.

Jason

On 2019/9/7 ??11:03, Jason Gunthorpe wrote:
> On Fri, Sep 06, 2019 at 06:02:35PM +0800, Jason Wang wrote:
>> On 2019/9/5 ??9:59, Jason Gunthorpe wrote:
>>> On Thu, Sep 05, 2019 at 08:27:34PM +0800, Jason Wang wrote:
>>>> Hi:
>>>>
>>>> Per request from Michael and Jason, the metadata accelreation
is
>>>> reverted in this version and rework in next version.
>>>>
>>>> Please review.
>>>>
>>>> Thanks
>>>>
>>>> Jason Wang (2):
>>>>     Revert "vhost: access vq metadata through kernel
virtual address"
>>>>     vhost: re-introducing metadata acceleration through kernel
virtual
>>>>       address
>>> There are a bunch of patches in the queue already that will help
>>> vhost, and I a working on one for next cycle that will help alot
more
>>> too.
>>
>> I will check those patches, but if you can give me some pointers or
keywords
>> it would be much appreciated.
> You can look here:
>
> https://github.com/jgunthorpe/linux/commits/mmu_notifier
>
> The first parts, the get/put are in the hmm tree, and the last part,
> the interval tree in the last commit is still a WIP, but it would
> remove alot of that code from vhost as well.
>
> Jason

Thanks a lot, will have a look at these and come back if I met any issues.

On 2019/9/6 ??9:15, David Miller wrote:
> From: Jason Wang <jasowang at redhat.com>
> Date: Fri, 6 Sep 2019 18:02:35 +0800
>
>> On 2019/9/5 ??9:59, Jason Gunthorpe wrote:
>>> I think you should apply the revert this cycle and rebase the other
>>> patch for next..
>>>
>>> Jason
>> Yes, the plan is to revert in this release cycle.
> Then you should reset patch #1 all by itself targetting 'net'.

Thanks for the reminding. I want the patch to go through Michael's vhost  
tree, that's why I don't put 'net' prefix. For next time, maybe
I can
use "vhost" as a prefix for classification?