[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20190117152142.GB20359@infradead.org>
Date: Thu, 17 Jan 2019 07:21:42 -0800
From: Christoph Hellwig <hch@...radead.org>
To: Vincent Whitchurch <vincent.whitchurch@...s.com>
Cc: Arnd Bergmann <arnd@...db.de>, sudeep.dutt@...el.com,
ashutosh.dixit@...el.com, gregkh <gregkh@...uxfoundation.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Kishon Vijay Abraham I <kishon@...com>,
Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
linux-pci <linux-pci@...r.kernel.org>,
linux-ntb@...glegroups.com, Jon Mason <jdmason@...zu.us>,
Dave Jiang <dave.jiang@...el.com>,
Allen Hubbe <allenbh@...il.com>
Subject: Re: [PATCH 0/8] Virtio-over-PCIe on non-MIC
On Thu, Jan 17, 2019 at 04:19:06PM +0100, Vincent Whitchurch wrote:
> On the RC, a vop-host-backend driver (PCI driver) sets up some shared
> memory area which the RC and the endpoint can use to communicate the
> location of the MIC device descriptors and other information such as the
> MSI address. It implements vop callbacks to allow the vop framework to
> obtain the address of the MIC descriptors and send/receive interrupts
> to/from the guest.
Why would we require any work on the RC / host side? A properly
setup software controlled virtio device should just show up as a
normal PCIe device, and the virtio-pci device should bind to it.
Powered by blists - more mailing lists