Re: [PATCH v15 00/25] RTRS (former IBTRS) RDMA Transport Library and RNBD (former IBNBD) RDMA Network Block Device

From: Danil Kipnis <danil.kipnis@cloud.ionos.com>
To: Jens Axboe <axboe@kernel.dk>
Cc: linux-block@vger.kernel.org, linux-rdma@vger.kernel.org,
	Jason Gunthorpe <jgg@ziepe.ca>,
	Bart Van Assche <bvanassche@acm.org>,
	Leon Romanovsky <leon@kernel.org>,
	Doug Ledford <dledford@redhat.com>,
	Jinpu Wang <jinpu.wang@cloud.ionos.com>
Subject: Re: [PATCH v15 00/25] RTRS (former IBTRS) RDMA Transport Library and RNBD (former IBNBD) RDMA Network Block Device
Date: Fri, 15 May 2020 16:42:12 +0200	[thread overview]
Message-ID: <CAHg0Hux7J3yieOiOxkqPOKNxj07T_hU-x2UF5Dt+uy9Kku4WsA@mail.gmail.com> (raw)
In-Reply-To: <e04fc798-ee25-53a5-fae0-5985306b55fd@kernel.dk>

On Fri, May 15, 2020 at 3:53 PM Jens Axboe <axboe@kernel.dk> wrote:
>
> On 5/15/20 4:29 AM, Danil Kipnis wrote:
> > Hi Jens,
> >
> > we've fixed the kbuild cross-compile problem identified for our
> > patches for 5.7-rc4. The block part has been reviewed by Bart van
> > Assche (thanks a lot Bart), we also replaced idr by xarray there as
> > Jason suggested. You planned to queue us
> > for 5.7: https://www.spinics.net/lists/linux-rdma/msg88472.html. Could
> > you please give Jason an OK to take this through the rdma tree, see
> > https://www.spinics.net/lists/linux-rdma/msg91400.html?
>
> My main worry isn't the current state of it, it's more how it's going
> to be handled going forward. If you're definitely going to maintain
> the upstream code in a suitable fashion, and not maintain an on-the-side
> version that you push to clients, then I'm fine with it going upstream
> and you can add my acked-by to the block part of the series.
>
> But maintaining the upstream version as the canonical version is key
> here.
Thanks a lot for your reply. We only do maintain the upstream code: we
have an extra compatibility layer which allows us to compile this code
for older kernels still in use in our production and we plan
to continue to do so in the future.
The only real difference of this code to the one running in our production
right now is the name of the driver: it is still "ibnbd" instead of current rnbd
but we will switch to it for the new kernels and also for the older kernels as
soon as our internal provisioning and monitoring infrastructure is prepared.

@Jason Should I resent you the patchset with Acked-By Jens added?

On Fri, May 15, 2020 at 3:53 PM Jens Axboe <axboe@kernel.dk> wrote:
>
> On 5/15/20 4:29 AM, Danil Kipnis wrote:
> > Hi Jens,
> >
> > we've fixed the kbuild cross-compile problem identified for our
> > patches for 5.7-rc4. The block part has been reviewed by Bart van
> > Assche (thanks a lot Bart), we also replaced idr by xarray there as
> > Jason suggested. You planned to queue us
> > for 5.7: https://www.spinics.net/lists/linux-rdma/msg88472.html. Could
> > you please give Jason an OK to take this through the rdma tree, see
> > https://www.spinics.net/lists/linux-rdma/msg91400.html?
>
> My main worry isn't the current state of it, it's more how it's going
> to be handled going forward. If you're definitely going to maintain
> the upstream code in a suitable fashion, and not maintain an on-the-side
> version that you push to clients, then I'm fine with it going upstream
> and you can add my acked-by to the block part of the series.
>
> But maintaining the upstream version as the canonical version is key
> here.
>
> --
> Jens Axboe
>