[libvirt] [PATCH] qemu: map "virtio" video model to "virt" machtype correctly (arm/aarch64)

[Laszlo Ersek]

On 09/16/16 08:28, Martin Kletzander wrote:
> On Fri, Sep 16, 2016 at 06:04:46AM +0200, Laszlo Ersek wrote:
>> Most of QEMU's PCI display device models, such as:
>>
>>  libvirt video/model/@type  QEMU -device
>>  -------------------------  ------------
>>  cirrus                     cirrus-vga
>>  vga                        VGA
>>  qxl                        qxl-vga
>>  virtio                     virtio-vga
>>
>> come with a linear framebuffer (sometimes called "VGA compatibility
>> framebuffer"). This linear framebuffer lives in one of the PCI device's
>> MMIO BARs, and allows guest code (primarily: firmware drivers, and
>> non-accelerated OS drivers) to display graphics with direct memory
>> access.
>>
>> Due to architectural reasons on aarch64/KVM hosts, this kind of
>> framebuffer doesn't / can't work in
>>
>>  qemu-system-(arm|aarch64) -M virt
>>
>> machines. Cache coherency issues guarantee a corrupted / unusable
>> display.
>> The problem has been researched by several people, including kvm-arm
>> maintainers, and it's been decided that the best way (practically the
>> only
>> way) to have boot time graphics for such guests is to consolidate on
>> QEMU's "virtio-gpu-pci" device.
>>
>>> From <https://bugzilla.redhat.com/show_bug.cgi?id=1195176>, libvirt
>> supports
>>
>>  <devices>
>>    <video>
>>      <model type='virtio'/>
>>    </video>
>>  </devices>
>>
>> but libvirt unconditionally maps @type='virtio' to QEMU's "virtio-vga"
>> device model. (See the qemuBuildDeviceVideoStr() function and the
>> "qemuDeviceVideo" enum impl.)
>>
>> According to the above, this is not right for the "virt" machine type;
>> the
>> qemu-system-(arm|aarch64) binaries don't even recognize the "virtio-vga"
>> device model (justifiedly). Whereas "virtio-gpu-pci", which is a pure
>> virtio device without a compatibility framebuffer, is available, and
>> works
>> fine.
>>
>> (The ArmVirtQemu ("AAVMF") platform of edk2 -- that is, the UEFI firmware
>> for "virt" -- supports "virtio-gpu-pci", as of upstream commit
>> 3ef3209d3028. See
>> <https://tianocore.acgmultimedia.com/show_bug.cgi?id=66>.)
>>
>> Override the default mapping of "virtio", from "virtio-vga" to
>> "virtio-gpu-pci", if qemuDomainMachineIsVirt() evaluates to true.
>>
>> Cc: Andrea Bolognani <abologna at redhat.com>
>> Cc: Drew Jones <drjones at redhat.com>
>> Cc: Marc-André Lureau <marcandre.lureau at redhat.com>
>> Suggested-by: Marc-André Lureau <marcandre.lureau at redhat.com>
>> Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1372901
>> Signed-off-by: Laszlo Ersek <lersek at redhat.com>
> 
> Very nicely written, makes sense and since virtio-vga didn't work for
> virt machines, there is no problem with changing it.

Thank you!

> One small nit though, syntax-check will complain that one of the lines
> in the .args file is longer, you can either use 'tests/test-wrap-argv.pl
> --in-place
> tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args'
> or just squash this diff in, whatever you find easier:

While working on the test data, I consulted other files, and I noticed
that they were carefully wrapped. I tried to do the same -- did I fail?

$ wc -L \
  tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args
74 tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args

and "tests/test-wrap-argv.pl" has

    my $max_len = 78;

Hmmm... After running the script, it looks like I was too cautious. The
problem is actually that I broke one of those lines too early, and now
it's too short. :)

I'll send a v2.

Thanks!
Laszlo

> diff --git
> i/tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args
> w/tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args
> index 56dbdfb66fa2..76ee977a3ca2 100644
> --- i/tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args
> +++ w/tests/qemuxml2argvdata/qemuxml2argv-aarch64-video-virtio-gpu-pci.args
> @@ -20,7 +20,7 @@ QEMU_AUDIO_DRV=none \
> addr=0x1 \
> -device ioh3420,port=0x9,chassis=2,id=pci.2,bus=pcie.0,multifunction=on,\
> addr=0x1.0x1 \
> --device virtio-net-pci,vlan=0,id=net0,mac=52:54:00:73:34:53,bus=pci.1,\
> -addr=0x0,bootindex=1 \
> +-device
> virtio-net-pci,vlan=0,id=net0,mac=52:54:00:73:34:53,bus=pci.1,addr=0x0,\
> +bootindex=1 \
> -net user,vlan=0,name=hostnet0 \
> -device virtio-gpu-pci,id=video0,bus=pci.2,addr=0x0
> -- 
> 
> Other than that, with that small thing fixed:
> 
> Acked-by: Martin Kletzander <mkletzan at redhat.com>