os-realtek-re: interface hangs after link flap in OPNsense 26.1.11_5 when used in multi-port bridge (ASM1806 PCIe switch)

[KimiyaMorozova]

Important notices
Before you add a new report, we ask you kindly to acknowledge the following:

• I have read the contributing guide lines at https://github.com/opnsense/plugins/blob/master/CONTRIBUTING.md
• I have searched the existing issues, open and closed, and I'm convinced that mine is new.
• The title contains the plugin to which this issue belongs

Describe the bug
2 Days ago everything was still working like normally. The Network Card using the if_re driver stop forwarding traffic after a physical link flap (cable disconnect/reconnect or endpoint link reset). The link state is still correctly reported as UP, but all traffic forwarding through the affected interface port stops working.

The system remains reachable via console or other ports, but network connectivity through the affected port becomes non-functional until the interface is manually reloaded (e.g. saving the interface configuration in the GUI or executing ifconfig down && ifconfig up).

To Reproduce
Steps to reproduce the behavior:

1. disconnect cable/shut down a pc
2. reconnect cable/start a pc
3. No traffic is forwarded to the opnsense machine

Expected behavior
After a physical link flap (cable disconnect/reconnect or endpoint restart), the interface and bridge should automatically recover without manual intervention. Traffic forwarding, routing, DHCP etc. should resume immediately once the link is back in an UP state. No manual interface reload or configuration save should be required.

Additional context
The issue is reproducible after any link flap event on the affected interfaces. The system reports correct link state transitions (DOWN → UP). However, forwarding through the interface stops entirely until the interface is manually reset.
All my Interfaces are behind a bridge.
Temporary workaround is always successful:

• saving interface configuration in the OPNsense GUI (triggers full reinitialization), or
• manually bouncing the interface via ifconfig down/up
• pressing 11 to reload all services

Environment

OPNsense 26.1.11_5-amd64
FreeBSD 14.3-RELEASE-p16
OpenSSL 3.0.21

RTL8125 2.5GbE (4-port PCIe card)
//Terminal Output//

re0@pci0:3:0:0: class=0x020000 rev=0x05 hdr=0x00 vendor=0x10ec device=0x8125 subvendor=0x10ec subdevice=0x0123
    vendor     = 'Realtek Semiconductor Co., Ltd.'
    device     = 'RTL8125 2.5GbE Controller'
    class      = network
    subclass   = ethernet
re1@pci0:4:0:0: class=0x020000 rev=0x05 hdr=0x00 vendor=0x10ec device=0x8125 subvendor=0x10ec subdevice=0x0123

# pciconf -lv | grep -B3 're0'
    device     = 'ASM1806 4-Port PCIe x2 Gen2 Packet Switch'
    class      = bridge
    subclass   = PCI-PCI
re0@pci0:3:0:0: class=0x020000 rev=0x05 hdr=0x00 vendor=0x10ec device=0x8125 subvendor=0x10ec subdevice=0x0123

[KimiyaMorozova]

After further testing, i weas able to observe the following:

• The issue affects all four Ports of my Quad Core RTL8125 PCIe Card interfaces (re0–re3).
• The integrated Realtek Gigabit interface (re4, WAN) is not affected.
• Hardware checksum offloading, TSO and LRO were disabled for testing without any change.
• A simple ifconfig reX and down ifconfig reX up immediately restores RX/TX operations on the effected port without restarting any services.

During the failure:
- interface reports "status: active"
- negotiated speed is correct
- tcpdump on the physical interface captures no packets
- interface statistics remain at zero