I try to run virtnet hotplug to make virtio for guest OS, made by VM. But it failed to make it.
$ sudo virtnet hotplug -i mlx5_0 -f 0x0 -m 0C:C4:7A:FF:22:93 -t 1500 -n 3 -s 1024
{'ib_device': 'mlx5_0', 'mac': '0C:C4:7A:FF:22:93', 'mtu': '0x5dc', 'num_queues': '0x3', 'max_queue_size': '0x400', 'features': '0x0', 'legacy': '0x0'}
{
"errno": -1071,
"errstr": "Failed to hotplug/unplug device, host is not ready"
}
Since firmware configuration setting is already set, it can’t make it.
$ sudo mlxconfig -d 03:00.0 -e q | grep -i \*
* NVME_EMULATION_NUM_PF 1 0 2
* PCI_SWITCH_EMULATION_ENABLE False(0) True(1) True(1)
* VIRTIO_BLK_EMULATION_NUM_MSIX 2 0 2
* VIRTIO_FS_EMULATION_NUM_MSIX 2 0 2
* VIRTIO_NET_EMULATION_ENABLE False(0) True(1) True(1)
* VIRTIO_NET_EMULATION_NUM_MSIX 2 0 10
* PER_PF_NUM_SF False(0) True(1) True(1)
* PF_BAR2_ENABLE True(1) False(0) False(0)
* SRIOV_EN True(1) False(0) False(0)
* NUM_VF_MSIX 11 0 11
* PF_TOTAL_SF 0 64 64
* PF_SF_BAR_SIZE 0 10 10
The '*' shows parameters with next value different from default/current value.
I wonder which parameter should be set at host side.
Currently, I set boot configuration as below.
intel_iommu=on iommu=pt intremap=no_x2apic_optout intel_pstate=disable default_hugepagesz=1g hugepagesz=1g hugepages=64 irqpoll pci=realloc