Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190116163253.23780-1-vincent.whitchurch@axis.com>
 <CAK8P3a2ifu9X+SXgoDPSFMYwnKr=Yex8_tD4qRimcOQAhuEa8Q@mail.gmail.com>
 <20190117105441.eqediwlekofp2srg@axis.com> <CAK8P3a2HtSJCLVdLcs=f0b5wv-k-6E_Z4Pt3Yp_kazUZh-Ek9A@mail.gmail.com>
 <20190117151529.GA3471@infradead.org> <20190117151920.GA20359@infradead.org>
In-Reply-To: <20190117151920.GA20359@infradead.org>
From:   Arnd Bergmann <arnd@arndb.de>
Date:   Thu, 17 Jan 2019 16:31:42 +0100
Message-ID: <CAK8P3a1r9ACPPuj90krYOB0yGVX3Usf86SYGxtnHpYCJ6_65hw@mail.gmail.com>
Subject: Re: [PATCH 0/8] Virtio-over-PCIe on non-MIC
To:     Christoph Hellwig <hch@infradead.org>
Cc:     Vincent Whitchurch <vincent.whitchurch@axis.com>,
        sudeep.dutt@intel.com, ashutosh.dixit@intel.com,
        gregkh <gregkh@linuxfoundation.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Kishon Vijay Abraham I <kishon@ti.com>,
        Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>,
        linux-pci <linux-pci@vger.kernel.org>,
        linux-ntb@googlegroups.com, Jon Mason <jdmason@kudzu.us>,
        Dave Jiang <dave.jiang@intel.com>,
        Allen Hubbe <allenbh@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Jan 17, 2019 at 4:19 PM Christoph Hellwig <hch@infradead.org> wrote:
>
> On Thu, Jan 17, 2019 at 07:15:29AM -0800, Christoph Hellwig wrote:
> > On Thu, Jan 17, 2019 at 01:39:27PM +0100, Arnd Bergmann wrote:
> > > Can you describe how you expect a VOP device over NTB or
> > > PCIe-endpoint would get created, configured and used?
> > > Is there always one master side that is responsible for creating
> > > virtio devices on it, with the slave side automatically attaching to
> > > them, or can either side create virtio devices? Is there any limit on
> > > the number of virtio devices or queues within a VOP device?
> >
> > For VOP device over NTB your configure your device using configfs
> > on one side, and for the other side it will just show up like any
> > other PCIe device, because it is.
>
> Sorry, I mean over the PCI-EP infratructure of course.  NTB actually
> is rather hairy and complicated.

My understanding was that with virtio, we would be able to have multiple
virtio devices on a single PCI-EP port, so you need a multi-step
configuration: You first set up the PCI-EP to instantiate a VOP device,
which is then seen on both ends of the connection. The question
is how to create a particular virtio device instance (or a set of those)
inside of it.

      Arnd