]> xenbits.xensource.com Git - libvirt.git/commit
conf: Add <hostdev model='virtio-{non-}transitional'/>
authorCole Robinson <crobinso@redhat.com>
Thu, 17 Jan 2019 17:52:33 +0000 (12:52 -0500)
committerCole Robinson <crobinso@redhat.com>
Mon, 4 Mar 2019 16:08:40 +0000 (11:08 -0500)
commitef41ff42197dbd3b252436c7302fd843ef7ec05c
treee4ff0bd90a9885303cf90cb592959ed3af3f946a
parent4d964373b51f47fbbc2df2a58bb01ea033660d36
conf: Add <hostdev model='virtio-{non-}transitional'/>

qemu vhost-scsi devices map to XML roughly like:

    <hostdev mode='subsystem' type='scsi_host'>
      <source protocol='vhost' wwpn=X/>
    </hostdev>

To support vhost-scsi-pci-{non-}traditional in qemu, we
need to to extend the SCSI Host hostdev XML to handle
model= value. This matches the XML model= format used
for mediated devices. This is just the domain_conf bits
and some XML test cases.

Use of virtio-X naming here does not match the hostdev
protocol=vhost nor does it match the qemu vhost-X device
naming, however it's more consistent with all other
model= names in this area, and also matches the
inconsistency of <vsock> devices which use model=virtio
but map to vhost-vsock on the qemu commandline

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
13 files changed:
docs/formatdomain.html.in
docs/schemas/domaincommon.rng
src/conf/domain_conf.c
src/conf/domain_conf.h
src/libvirt_private.syms
tests/qemuxml2argvdata/virtio-non-transitional.x86_64-3.1.0.args
tests/qemuxml2argvdata/virtio-non-transitional.x86_64-latest.args
tests/qemuxml2argvdata/virtio-non-transitional.xml
tests/qemuxml2argvdata/virtio-transitional.x86_64-3.1.0.args
tests/qemuxml2argvdata/virtio-transitional.x86_64-latest.args
tests/qemuxml2argvdata/virtio-transitional.xml
tests/qemuxml2xmloutdata/virtio-non-transitional.xml
tests/qemuxml2xmloutdata/virtio-transitional.xml