"Daniel P. Berrange" <berrange@xxxxxxxxxx> writes: > From the XML pov, we shouldn't restrict use of <model> to only be > for FS_TYPE_MOUNT. That is a QEMU driver implementation restriction, > so just make the QEMU driver raise VIR_ERR_CONFIG_UNSUPPORTED for > the cases we don't want, when building the CLI args OK to amend this piece? The QEMU driver supports only FS_TYPE_MOUNT so the additional check seems reduntant there. diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 98dbe14..c897f1b 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -15806,10 +15806,6 @@ virDomainFSDefFormat(virBufferPtr buf, case VIR_DOMAIN_FS_TYPE_MOUNT: virBufferEscapeString(buf, "<source dir='%s'/>\n", def->src); - if (def->model) { - virBufferEscapeString(buf, "<model type='%s'/>\n", - virDomainFSModelTypeToString(def->model)); - } break; case VIR_DOMAIN_FS_TYPE_BIND: @@ -15838,6 +15834,11 @@ virDomainFSDefFormat(virBufferPtr buf, break; } + if (def->model) { + virBufferEscapeString(buf, "<model type='%s'/>\n", + virDomainFSModelTypeToString(def->model)); + } + virBufferEscapeString(buf, "<target dir='%s'/>\n", def->dst); Thanks, Giuseppe -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list