Two reasons:
1.in none hotplug, we will pass it. We can see from libvirt function
qemuBuildVhostuserCommandLine
2.qemu will use this vetcor num to init msix table. If we don't pass, qemu
will use default value, this will cause VM can only use default value
interrupts at most.
Signed-off-by: gaohaifeng <gaohaifeng.gao(a)huawei.com>
---
src/qemu/qemu_hotplug.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c
index 7c3f32d..55af888 100644
--- a/src/qemu/qemu_hotplug.c
+++ b/src/qemu/qemu_hotplug.c
@@ -946,6 +946,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver,
char **vhostfdName = NULL;
int *vhostfd = NULL;
size_t vhostfdSize = 0;
+ size_t queueSize = 0;
char *nicstr = NULL;
char *netstr = NULL;
virNetDevVPortProfilePtr vport = NULL;
@@ -1010,6 +1011,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver,
tapfdSize = vhostfdSize = net->driver.virtio.queues;
if (!tapfdSize)
tapfdSize = vhostfdSize = 1;
+ queueSize = tapfdSize;
if (VIR_ALLOC_N(tapfd, tapfdSize) < 0)
goto cleanup;
memset(tapfd, -1, sizeof(*tapfd) * tapfdSize);
@@ -1029,6 +1031,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver,
tapfdSize = vhostfdSize = net->driver.virtio.queues;
if (!tapfdSize)
tapfdSize = vhostfdSize = 1;
+ queueSize = tapfdSize;
if (VIR_ALLOC_N(tapfd, tapfdSize) < 0)
goto cleanup;
memset(tapfd, -1, sizeof(*tapfd) * tapfdSize);
@@ -1049,6 +1052,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver,
tapfdSize = vhostfdSize = net->driver.virtio.queues;
if (!tapfdSize)
tapfdSize = vhostfdSize = 1;
+ queueSize = tapfdSize;
if (VIR_ALLOC_N(tapfd, tapfdSize) < 0)
goto cleanup;
memset(tapfd, -1, sizeof(*tapfd) * tapfdSize);
@@ -1079,6 +1083,9 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver,
break;
case VIR_DOMAIN_NET_TYPE_VHOSTUSER:
+ queueSize = net->driver.virtio.queues;
+ if (!queueSize)
+ queueSize = 1;
if (!qemuDomainSupportsNetdev(vm->def, priv->qemuCaps, net)) {
virReportError(VIR_ERR_INTERNAL_ERROR,
"%s", _("Netdev support unavailable"));
@@ -1223,7 +1230,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver,
VIR_FORCE_CLOSE(vhostfd[i]);
if (!(nicstr = qemuBuildNicDevStr(vm->def, net, vlan, 0,
- vhostfdSize, priv->qemuCaps)))
+ queueSize, priv->qemuCaps)))
goto try_remove;
qemuDomainObjEnterMonitor(driver, vm);
--
2.6.4
Show replies by date
On 01.12.2016 07:01, gaohaifeng wrote:
Two reasons:
1.in none hotplug, we will pass it. We can see from libvirt function
qemuBuildVhostuserCommandLine
2.qemu will use this vetcor num to init msix table. If we don't pass, qemu
will use default value, this will cause VM can only use default value
interrupts at most.
Signed-off-by: gaohaifeng <gaohaifeng.gao(a)huawei.com>
---
src/qemu/qemu_hotplug.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
ACKed and pushed.
Michal