Hi Ranjit.

I’m opening up this conversation to everyone by cc’ing the spdk mail list.

Yes, the upcoming changes we have will complete the separation of dpdk and spdk.  

They will also abstract all of the posix dependencies (e.g. pthread_create becomes spdk_thread_create).

The question about how many components to refactor needs to be agreed upon.  

At this point the feedback I’ve received from Ben and Daniel has been that they don’t want to convert all components at once.  So I’ve been taking my pull requests apart into smaller pieces.  Here is one of my older pull requests for reference.


This can be compared with my current pull request:

SPDK EAL Abstractions

and with yours.

Abstract and move rte_ functions into spdk functions that act ad pass-through functions in env_dpdk

In terms of timeline, I’d like to complete these changes ASAP.  If we can’t merge everything we have into master relatively soon (like this week) then I’d like to propose that a separate SPDK "platform_abstraction" branch be created (not a Github fork) so all of the environmental changes can be submitted there.  This will allow contributors to develop these changes away from master more quickly, and the process of merging SHAs from the “platform_abstraction” into master can go more slowly.

As was discussed at the Intel Builders conference last week, we should set up a periodic conference call to others in the community can discuss this and other subjects.   I’ll work with Jim, Ben, and Daniel to set that up.

John Meneghini

On Apr 26, 2017, at 2:30 AM, Ranjit Noronha <notifications@github.com> wrote:

Hi John,

Sure, lets work together on this.

Will your upcoming changes complete the separation of dpdk and spdk? Or
will there be components still left to refactor?

Do you have a ballpark timeline on when these commits will complete?



On 4/26/2017 3:39 AM, John Meneghini wrote:
> *@johnmeneghini* commented on this pull request.
> Hi Ranjit.
> Some of these changes conflict with the changes in my pull requests
> at: #136
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_spdk_spdk_pull_136&d=DwMCaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=YwY3G2IQGMmRKLMucJawWMY6oKqgVtsYGUHz2coslFU&m=R27IwlD_OQfN2LQ2hkEpFRkqQUUFbT5cYejmHd7bMvU&s=0IhEuLlRUhA5FtGYPBm2mvudeMq_-Kc1VhrTHtiSQpw&e=>
> #148
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_spdk_spdk_pull_148&d=DwMCaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=YwY3G2IQGMmRKLMucJawWMY6oKqgVtsYGUHz2coslFU&m=R27IwlD_OQfN2LQ2hkEpFRkqQUUFbT5cYejmHd7bMvU&s=ZnpO3TMYPeag6INMpJ5vy85vjCi2DKg-yevaO32OKdI&e=>
> #150
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_spdk_spdk_pull_150&d=DwMCaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=YwY3G2IQGMmRKLMucJawWMY6oKqgVtsYGUHz2coslFU&m=R27IwlD_OQfN2LQ2hkEpFRkqQUUFbT5cYejmHd7bMvU&s=uE3orXmf-gHcfwhisC8V06GzJskm3PgJZiXo-blzzGM&e=>
> As you can see, I've refactored my pull request several times, and I
> have many more commits waiting to go in.
> I spoke to Ben about this last week and I think our changes
> percipitated Ben's email on the reflector. Let's see if we can work
> together to get these changes in.
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_spdk_spdk_pull_142-23pullrequestreview-2D34702507&d=DwMCaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=YwY3G2IQGMmRKLMucJawWMY6oKqgVtsYGUHz2coslFU&m=R27IwlD_OQfN2LQ2hkEpFRkqQUUFbT5cYejmHd7bMvU&s=f41NFqdxN9qUse424Fu6Q6nzgg47hYCNIpUqgLTi1jM&e=>,
> or mute the thread
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AaZUlq2LJd6qHZxPs8-5Ff5wNL-5FbMB-2DH5Iks5rzm8HgaJpZM4M-5FI8T&d=DwMCaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=YwY3G2IQGMmRKLMucJawWMY6oKqgVtsYGUHz2coslFU&m=R27IwlD_OQfN2LQ2hkEpFRkqQUUFbT5cYejmHd7bMvU&s=AdN-T6Jwj2Q3QPuLK2TFJV5Ba94utbSFH8vBd2j6QPo&e=>.

You are receiving this because you commented.
Reply to this email directly, view it on GitHub, or mute the thread.