Hi

On Mon, May 3, 2021 at 4:07 PM Firecode95 <1926952@bugs.launchpad.net>
wrote:

> Also just removing spice part makes it bootable:
> -spice port=5900,addr=127.0.0.1,disable-ticketing=on
>
> --
> You received this bug notification because you are a member of qemu-
> devel-ml, which is subscribed to QEMU.
> https://bugs.launchpad.net/bugs/1926952
>
> Title:
>   SPICE support broken with 6.0
>
> Status in QEMU:
>   New
>
> Bug description:
>   Using latest relase 6.0.0 while using Intel GVT-G DMA-BUF and SPICE
>   for usb redirection Qemu won't start:
>
>   qemu-system-x86_64: The console requires display DMABUF support.
>
>   However just patching ui/console.c:
>
>   if (flags & GRAPHIC_FLAGS_DMABUF &&
>           !displaychangelistener_has_dmabuf(dcl)) {
>           error_setg(errp, "The console requires display DMABUF support.");
>           return false;
>   }
>
>   to always return true for dmabuf part works just fine:
>
>   if (flags & GRAPHIC_FLAGS_DMABUF &&
>           !displaychangelistener_has_dmabuf(dcl)) {
>           error_setg(errp, "The console requires display DMABUF support.");
>           return true;
>   }
>
>   This behavior wasn't in qemu 5.x version.
>
>   To reproduce this bug need to use:
>
>   /usr/bin/qemu-system-x86_64 \
>   -machine q35 \
>   -enable-kvm \
>   -no-user-config \
>   -nodefaults \
>   -no-hpet \
>   -display gtk,gl=on \
>   -device
> pcie-root-port,port=0x0,chassis=1,id=pci.1,bus=pcie.0,multifunction=on,addr=0x1
> \
>   -device
> vfio-pci,id=hostdev2,driver=vfio-pci-nohotplug,romfile=/sys/devices/pci0000:00/0000:00:02.0/gvt_firmware,sysfsdev=/sys/bus/mdev/devices/1ae40c36-b180-4af0-8fab-c27de21f597d,x-igd-opregion=on,ramfb=on,display=on,xres=1920,yres=1080,bus=pcie.0,multifunction=on,addr=0x2
> \
>   -spice port=5900,addr=127.0.0.1,disable-ticketing=on
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/qemu/+bug/1926952/+subscriptions
>
>

Did you actually get the dmabuf update displayed over spice? If not, then
it is not a bug, it's a bug fix :) But I might be missing something.. I
don't see how this could happen without egl-headless though.


-- 
Marc-André Lureau