This series adds support for a vgpu display to the qemu vfio code.
For now only regions are supported, dmabufs will follow later.

The vfio API update is done, queued in drm-next, should land in the
upstream kernel during the 4.16 merge window.  So the 4.16-rc1 kernel
header sync should bring the header changes needed for this series.

Patch #1 of this series has the vfio.h updates too, for testing
convinience, but I don't plan to include that patch in the final
patch submission.

plese test and review,
  Gerd

Gerd Hoffmann (5):
  headers: update linux-headers/linux/vfio.h (intel-gvt kernel patches,
    v17)
  headers: add drm/drm_fourcc.h to standard-headers
  ui/pixman: add qemu_drm_format_to_pixman()
  vfio/display: core & wireup
  vfio/display: adding region support

 hw/vfio/pci.h                             |   4 +
 include/hw/vfio/vfio-common.h             |   8 +
 include/standard-headers/drm/drm_fourcc.h | 382 ++++++++++++++++++++++++++++++
 include/ui/qemu-pixman.h                  |   5 +
 linux-headers/linux/vfio.h                |  64 +++++
 hw/vfio/display.c                         | 149 ++++++++++++
 hw/vfio/pci.c                             |   9 +
 ui/qemu-pixman.c                          |  22 ++
 hw/vfio/Makefile.objs                     |   2 +-
 scripts/update-linux-headers.sh           |   4 +
 10 files changed, 648 insertions(+), 1 deletion(-)
 create mode 100644 include/standard-headers/drm/drm_fourcc.h
 create mode 100644 hw/vfio/display.c

-- 
2.9.3

On Wed, 31 Jan 2018 13:12:12 +0100
Gerd Hoffmann <kraxel@redhat.com> wrote:

> This series adds support for a vgpu display to the qemu vfio code.
> For now only regions are supported, dmabufs will follow later.
> 
> The vfio API update is done, queued in drm-next, should land in the
> upstream kernel during the 4.16 merge window.  So the 4.16-rc1 kernel
> header sync should bring the header changes needed for this series.
> 
> Patch #1 of this series has the vfio.h updates too, for testing
> convinience, but I don't plan to include that patch in the final
> patch submission.
> 
> plese test and review,

Hi Gerd,

The vfio bits look reasonable to me, it'd be nice if we could operate
on the VFIODevice rather than VFIOPCIDevice to make this universal for
all vfio devices, but that's just a nit since non-PCI graphics devices
may never come to fruition.  As for testing, region support was
included for NVIDIA, dmabuf for Intel... is this testing request mainly
for the Kirti and others at NVIDIA?  Thanks,

Alex

  Hi,

> > plese test and review,
> 
> may never come to fruition.  As for testing, region support was
> included for NVIDIA, dmabuf for Intel... is this testing request mainly
> for the Kirti and others at NVIDIA?  Thanks,

Yes.  v1 was tested by Kirti and I've fixed up two small issues pointed
out, so this should work as-is in theory.  But confirming that it
actually works before merging would be great.

thanks,
  Gerd

On 2/1/2018 1:54 PM, Gerd Hoffmann wrote:
>   Hi,
> 
>>> plese test and review,
>>
>> may never come to fruition.  As for testing, region support was
>> included for NVIDIA, dmabuf for Intel... is this testing request mainly
>> for the Kirti and others at NVIDIA?  Thanks,
> 
> Yes.  v1 was tested by Kirti and I've fixed up two small issues pointed
> out, so this should work as-is in theory.  But confirming that it
> actually works before merging would be great.
> 

Thanks Gerd for taking care of the minor fixes.

Tested this series for region support.

This patch-set looks good to me.

Tested-by: Kirti Wankhede <kwankhede@nvidia.com>

Thanks,
Kirti