Re: [PATCH v4 1/2] fpga: fpga-mgr: Add readback support

[Thor Thayer]

Hi Moritz,

On 10/7/19 4:20 PM, Moritz Fischer wrote:
> Hi Thor,
>
> On Mon, Oct 07, 2019 at 01:06:51PM -0500, Thor Thayer wrote:
> > Hi Moritz,
> >
> > On 9/27/19 1:23 PM, Moritz Fischer wrote:
> > > Thor,
> > >
> > > On Fri, Sep 27, 2019 at 09:32:11AM -0500, Thor Thayer wrote:
> > > > Hi Kedar & Moritz,
> > > >
> > > > On 9/27/19 12:13 AM, Appana Durga Kedareswara Rao wrote:
> > > > > Hi Alan,
> > > > >
> > > > > Did you get a chance to send your framework changes to upstream?
> > > No they weren't upstreamed.
> > >
> > > > > @Moritz Fischer: If Alan couldn't send his patch series, Can we take this patch series??
> > > > > Please let me know your thoughts on this.
> > >
> > > Alan had some comments RE: #defines, I'll have to take another look.
> > > > > Regards,
> > > > > Kedar.
> > > >
> > > >
> > > > I'd like to see some mechanism added as well. Our CvP driver needs a way to
> > > > load images to the FPGA over the PCIe bus.
> > >
> > > Can you elaborate a bit on the CvP use-case and how that would work? Who
> > > would use the device how after loading the bitstream?
> > >
> > > Generally there are several use cases that I have collected mentally
> > > over the years:
> > >
> > > I) DFL use case:
> > >     - Mixed-set of drivers: Kernel and Userspace
> > >     - FPGA logic is discoverable through DFL
> > >     - Userspace application wants to reprogram FPGA
> > >
> > > II) DT configfs use case:
> > >     - Mixed-set of drivers: Kernel and Userspace
> > >     - FPGA logic is *not* discoverable (hence DT overlay)
> > >     - Userspace application wants to reprogram FPGA
> > >
> > > III) Thomas' case:
> > >     - Kernel only drivers (pcie bridge, pcie drivers, ...)
> > >     - FPGA logic is fully discoverable (i.e. PCIe endpoint
> > >       implemented in FPGA, connected to SoC via PCIe)
> > >     - Userspace application wants to reprogram FPGA
> > >
> > > IV) VFIO case:
> > >     - Usually exposes either entire device via vfio-pci or part via
> > >       vfio-mdev
> > >     - Loading (basic) bitstream at boot from flash
> > >     - vfio-mdev case can use FPGA region interface + ioctl
> > >     - Full VFIO case is similar to III)
> > >
> > > How does your CvP use case fit in? Collecting all the use-cases would
> > > help with moving forward on coming up with an API :)
> > The CvP case is the same as III) Thomas' case. The FPGA configuration
> > bitstream is downloaded over the PCIe.
> >
> > The one difference in my case is that there isn't an SoC. This is a Intel
> > host processor connecting to a non-SoC Stratix10/Arria10. The non-SoC
> > A10/S10, boots a minimal image (CvP) setting up the peripheral pins and
> > enabling the PCIe endpoint for CvP downloads.
> >
> > The host can then download bitstreams using the FPGA Manager through debugFS
> > and when the bitstream finishes downloading and the FPGA enters User Mode,
> > the functionality is available for the host to use.
>
> I am generally confused by this driver. How does it work exactly? What
> happens after altera-cvp binds a PCI device?
>
> You can use it to download a bitstream (say we had the debugfs
> interface), and then what happens next? How do I use the device? It
> already has a PCI driver bound to it at that point?

Sorry for the delay. In the CvP case, I reboot the host device leaving 
the FPGA board powered so the new PCI interface is enumerated.

There may be a better way. I'll need to research Thomas' use case. There 
may be some good lessons to learn there.

Thanks,

Thor

> What happens next?
>
> Please tell me that not the only use-case for this is /dev/mem :)
>
> Thomas' use-case is different in that behind the FPGA device there are
> actual other *discoverable* PCI devices that will get enumerated and
> bind to separate drivers.
>
> Thanks,
> Moritz
>
> PS: I'll be out this week on vacation starting tmr so responses might be delayed