What were you trying to do that didn't work?

After starting vm with <link state='down'/> for e1000e or igb model type interface, login vm, get "Link detected: yes" and can ping outside.

Please provide the package NVR for which bug is seen:

libvirt-10.0.0-1.el9.x86_64
qemu-kvm-8.2.0-2.el9.x86_64

How reproducible:

100%

Steps to reproduce

1. Start vm with:

    <interface type='network'>
      <mac address='52:54:00:0d:34:ce'/>
      <source network='default'/>
      <model type='igb'/>
      <link state='down'/>
      <address type='pci' domain='0x0000' bus='0x01' slot='0x00' function='0x0'/>
    </interface>

1. virsh start yqz-test
   Domain 'yqz-test' started

1. ps aux|grep qemu-kvm|grep yqz-test|grep net

...-netdev {"type":"tap","fd":"24","id":"hostnet0"} -device

...

1. Check vm interface on host:

1. virsh domif-getlink yqz-test vnet9
   vnet9 down

1. Check vm interface in vm:

1. virsh console yqz-test

[root@localhost ~]# ethtool enp1s0|grep Link
Link detected: yes

[root@localhost ~]# ping www.baidu.com -c2
PING www.wshifen.com (45.113.192.101) 56(84) bytes of data.
64 bytes from 45.113.192.101 (45.113.192.101): icmp_seq=1 ttl=49 time=69.9 ms
64 bytes from 45.113.192.101 (45.113.192.101): icmp_seq=2 ttl=49 time=70.0 ms

Expected results

Starting vm with <link state='down'/> for e1000e or igb model interface but get an actually up link in vm.

Actual results

The link in vm should also be down.

Additional info:

1. If we do 'virsh domif-setlink xxx up' then 'virsh domif-setlink xxx down', we will get expected results: "Link detected: no" and ping doesn't work.

2. Not reproduced for model type 'virtio' or 'rtl8139'.

3. Not reproduced by pure qemu way of  RHEL-1218.