On 11/23/2010 03:50 PM, Eric Blake wrote:
On 11/18/2010 08:31 AM, Stefan Berger wrote:
[apologies for letting this delay a week]
This patch introduces the usage of the pre-associate state of the IEEE 802.1Qbg standard on incoming VM migration on the target host. It is in response to bugzilla entry 632750.
https://bugzilla.redhat.com/show_bug.cgi?id=632750
For being able to differentiate the exact reason as to why a macvtap device is being created, either due to a VM creation or an incoming VM migration, I needed to pass that reason as a parameter from wherever qemudStartVMDaemon is being called in order to determine whether to send an ASSOCIATE (VM creation) or a PRE-ASSOCIATE (incoming VM migration) towards lldpad.
I am also fixing problem with the virsh domainxml-to-native call on the way.
Gerhard successfully tested the patch with a recent blade network 802.1Qbg-compliant switch.
The patch should not have any side-effects on the 802.1Qbh support in libvirt, but Scott (cc'ed) may want to verify this.
Signed-off-by: Stefan Berger<stefanb@us.ibm.com> Signed-off-by: Gerhard Stenzel<gerhard.stenzel@de.ibm.com> @@ -3951,7 +3953,8 @@ int qemudBuildCommandLine(virConnectPtr int **vmfds, int *nvmfds, const char *migrateFrom, - virDomainSnapshotObjPtr current_snapshot) + virDomainSnapshotObjPtr current_snapshot, + enum vmOperation vmop) Oh joy - a merge conflict with my virCommand stuff. If you push first, then I get to pick up the pieces :)
@@ -11844,6 +11851,41 @@ cleanup: return ret; }
+static void +qemudVPAssociatePortProfiles(virDomainDefPtr def) { + int i; + int last_good_net = -1; + virDomainNetDefPtr net; + + for (i = 0; i< def->nnets; i++) { + net = def->nets[i]; + if (net->type == VIR_DOMAIN_NET_TYPE_DIRECT) { + if (vpAssociatePortProfileId(net->ifname, + net->mac, + net->data.direct.linkdev, +&net->data.direct.virtPortProfile, + def->uuid, + VM_OP_MIGRATE_IN_FINISH) != 0) + goto err_exit; + } + last_good_net = i; + } + + return; + +err_exit: + for (i = 0; i< last_good_net; i++) { Should this loop be from (last_good_net,0] instead of [0,last_good_net) (that is, inverse of creation order)?
The order of the teardown doesn't matter from what I can see.
Index: libvirt-acl/src/util/macvtap.c =================================================================== --- libvirt-acl.orig/src/util/macvtap.c +++ libvirt-acl/src/util/macvtap.c @@ -77,9 +77,22 @@ # define LLDPAD_PID_FILE "/var/run/lldpad.pid"
+static const char *vmOpStr[] = { + [VM_OP_CREATE] = "create", + [VM_OP_SAVE] = "save", + [VM_OP_RESTORE] = "restore", + [VM_OP_DESTROY] = "destroy", + [VM_OP_MIGRATE_OUT] = "migrate out", + [VM_OP_MIGRATE_IN_START] = "migrate in start", + [VM_OP_MIGRATE_IN_FINISH] = "migrate in finish", + [VM_OP_NO_OP] = "no-op" This seems like it would better be done via VIR_ENUM_DECL and VIR_ENUM_IMPL?
Yes, I'll convert it and repost. Stefan