Hi
From: Gurchetan Singh <gurchetansingh@google.com>
v3 of "virtio-gpu cleanups and obvious definitions"
https://lists.gnu.org/archive/html/qemu-devel/2023-04/msg05392.html
All patches have been reviewed, though there was a question from
Bernhard Beschow about patch (3) and how it fits with the QOM:
https://lists.gnu.org/archive/html/qemu-devel/2023-05/msg00057.html
I go into detail in patch 3 commit message, but I think we meet
the requirements (which are tricky/fuzzy anyways). Also, I think
this is the cleanest way to add another 3D virtgpu backend. But
if anyone has other ideas, please do reply/review.
Antonio Caggiano (1):
virtio-gpu: CONTEXT_INIT feature
Dr. David Alan Gilbert (1):
virtio: Add shared memory capability
Gurchetan Singh (3):
hw/display/virtio-gpu-virgl: virtio_gpu_gl -> virtio_gpu_virgl
hw/display/virtio-gpu-virgl: make GL device more library agnostic
hw/display/virtio-gpu-virgl: define callbacks in realize function
hw/display/virtio-gpu-base.c | 3 +
hw/display/virtio-gpu-gl.c | 114 +--------------------------
hw/display/virtio-gpu-virgl.c | 137 +++++++++++++++++++++++++++++++--
hw/virtio/virtio-pci.c | 18 +++++
include/hw/virtio/virtio-gpu.h | 11 +--
include/hw/virtio/virtio-pci.h | 4 +
6 files changed, 160 insertions(+), 127 deletions(-)
--
2.40.1.521.gf1e218fcd8-goog
This looks fine to me:
however, do you have a series rebased on top that makes use of those changes? (I think we may want to delay merging this one until it's actually needed)
Got it. Bernhard actually recommended a separate virtio-gpu-rutabaga device instead of virtio-gpu-gl. I went with that approach:
I'll send out a new full-featured series once testing/hack cleanups are complete -- so no further action is needed on your part.