[PATCH v2 RFC] nvme: improve performance for virtual NVMe devices

Mike Waychison mikew at google.com
Thu May 12 10:29:42 PDT 2016


On Thu, May 12, 2016 at 7:40 AM, Monish Shah <monish at google.com> wrote:
>
> [+mikew]
>
> Hello all,
>
> I don't have all the context, but this sounds like the same thing that Mike Waychison discussed with me recently.  Mike has been added to the Technical Working Group e-mail alias and can communicate on Google's behalf.  I'm happy to help, where I can.
>
> Monish
>
>
> Monish Shah | Hardware Engineer | monish at google.com | 650-253-2822
>
>
> On Thu, May 12, 2016 at 7:21 AM, Rob Nelson <rlnelson at google.com> wrote:
>>
>> Apologies for my absence here but I've since left my position working on virtualization here at Google.
>>
>> Monish is the lead contact for Google with the NVMe working group, so he might help here. However, Frank/Eric
>> are still on the virtualization team and can comment if they are interested in moving forward here.
>>
>> RIght after I left (end of last year), I thought there was already a proposal that was virtualization friendly
>> (not my extenstion) but something similar to reduce number of guest exits. I'll let Monish comment on
>> that since I have no clue where it went.
>>
>> On Thu, May 12, 2016 at 12:05 AM, Christoph Hellwig <hch at infradead.org> wrote:
>>>
>>> On Thu, May 12, 2016 at 01:05:32AM -0300, Helen Koike wrote:
>>> > No, I am not, that is why I am asking. If it's not a burden to you I would
>>> > appreciate some help to start the proposal with the NVMe technical
>>> > workgroup, otherwise I'll need to go through a more bureaucratic path to do
>>> > it in Google's name.
>>>
>>> As far as I understand the disclosure rules prohibit memebers from
>>> bringing up proposal from others.  It would be really helpful
>>> if you/Google were part of this.
>>
>>

-html

Hi Christoph,

Helen and team have been looking at this on my team's behalf.  In
summary, we'd like to work together to get the feature in question
accepted upstream, and understand that this needs to go through the
working group for it to be adopted.

I'm relatively unfamiliar with affecting change through the working
group so far, but would like to work together to do so nevertheless if
this is what is required.  Please help direct what is needed from us
to move this forward and we will do our best to oblige.

Thanks,

Mike Waychison



More information about the Linux-nvme mailing list