On 01.01.2017 17:24, Roman Bogorodskiy wrote:
Michal Privoznik wrote:
> After c07d1c1c4f got merged it uncovered couple of broken domain
> XMLs for bhyvexml2argv test. Some disk drives had incompatible
> type of address configured.
>
> Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com>
> ---
> tests/bhyvexml2argvdata/bhyvexml2argv-acpiapic.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-base.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-bhyveload-bootorder.xml | 4 ++--
> tests/bhyvexml2argvdata/bhyvexml2argv-bhyveload-bootorder1.xml | 4 ++--
> tests/bhyvexml2argvdata/bhyvexml2argv-bhyveload-bootorder2.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-bhyveload-bootorder3.xml | 4 ++--
> tests/bhyvexml2argvdata/bhyvexml2argv-bhyveload-bootorder4.xml | 4 ++--
> tests/bhyvexml2argvdata/bhyvexml2argv-bhyveload-explicitargs.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-console.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-custom-loader.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-disk-cdrom-grub.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-disk-cdrom.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-grub-bootorder.args | 4 ++--
> tests/bhyvexml2argvdata/bhyvexml2argv-grub-bootorder.xml | 6 +++---
> tests/bhyvexml2argvdata/bhyvexml2argv-grub-bootorder2.args | 4 ++--
> tests/bhyvexml2argvdata/bhyvexml2argv-grub-bootorder2.xml | 6 +++---
> tests/bhyvexml2argvdata/bhyvexml2argv-grub-defaults.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-localtime.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-macaddr.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-serial-grub-nocons.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-serial-grub.xml | 2 +-
> tests/bhyvexml2argvdata/bhyvexml2argv-serial.xml | 2 +-
> 22 files changed, 32 insertions(+), 32 deletions(-)
>
> diff --git a/tests/bhyvexml2argvdata/bhyvexml2argv-acpiapic.xml
b/tests/bhyvexml2argvdata/bhyvexml2argv-acpiapic.xml
> index 2be970e05..316aaf895 100644
> --- a/tests/bhyvexml2argvdata/bhyvexml2argv-acpiapic.xml
> +++ b/tests/bhyvexml2argvdata/bhyvexml2argv-acpiapic.xml
> @@ -15,7 +15,7 @@
> <driver name='file' type='raw'/>
> <source file='/tmp/freebsd.img'/>
> <target dev='hda' bus='sata'/>
> - <address type='pci' domain='0x0000' bus='0x00'
slot='0x02' function='0x0'/>
> + <address type='drive' controller='0' bus='0'
target='2' unit='0'/>
> </disk>
> <interface type='bridge'>
> <model type='virtio'/>
Yeah, the tests are now broken for bhyve indeed. However, there's a more
serious problem than tests. The current bhyve address allocation code
doesn't really care about 'drive' address type and uses pci addresses
even for hard disk drivers. It's implemented this way to mimic the bhyve
behaviour at the time of when this code was written, i.e.:
-s 2:0,ahci-cd,/tmp/cdrom.iso bhyve
Where "2" is PCI slot and "0" is the function. So I guess the main
thing
here is to adjust the address allocation code which I hoooope to do
during this winter holidays...
Ah, didn't realize this problem. However, I still think this should go
in as it also fixes bhyvexml2argvtest which is failing now. The problem
you are mentioning can be addressed later.
Michal