tg3 link state detection

tg3.c:v3.8 (July 14, 2004)
ACPI: PCI interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 185
eth0: Tigon3 [partno(BCM95704A6) rev 2002 PHY(5704)] (PCIX:133MHz:64-bit) 10/100/1000BaseT Ethernet 00:0d:56:fe:4f:e7
eth0: HostTXDS[1] RXcsums[1] LinkChgREG[1] MIirq[1] ASF[0] Split[0] WireSpeed[1] TSOcap[1]
ACPI: PCI interrupt 0000:02:00.1[B] -> GSI 17 (level, low) -> IRQ 193
eth1: Tigon3 [partno(BCM95704A6) rev 2002 PHY(5704)] (PCIX:133MHz:64-bit) 10/100/1000BaseT Ethernet 00:0d:56:fe:4f:e8
eth1: HostTXDS[1] RXcsums[1] LinkChgREG[1] MIirq[1] ASF[0] Split[0] WireSpeed[1] TSOcap[1]

With 2.6.8(.1)+the add_pin_to_irq fix we're seeing problem with link state detection using tg3 devices. The network in question has a severe case of cable monkeys, so the link is a bit jumpy.

From time to time the tg3's do not detect that the link is up again,
and thus has to be rebooted (ifdown/up has not been tested due to triggerhappy non-technical on-site monkeys.)

Is this a known problem?

Quoted 7 lines

tg3.c:v3.8 (July 14, 2004)
ACPI: PCI interrupt 0000:02:00.0[A] -> GSI 16 (level, low) -> IRQ 185
eth0: Tigon3 [partno(BCM95704A6) rev 2002 PHY(5704)] (PCIX:133MHz:64-bit) 10/100/1000BaseT Ethernet 00:0d:56:fe:4f:e7
eth0: HostTXDS[1] RXcsums[1] LinkChgREG[1] MIirq[1] ASF[0] Split[0] WireSpeed[1] TSOcap[1]
ACPI: PCI interrupt 0000:02:00.1[B] -> GSI 17 (level, low) -> IRQ 193
eth1: Tigon3 [partno(BCM95704A6) rev 2002 PHY(5704)] (PCIX:133MHz:64-bit) 10/100/1000BaseT Ethernet 00:0d:56:fe:4f:e8
eth1: HostTXDS[1] RXcsums[1] LinkChgREG[1] MIirq[1] ASF[0] Split[0] WireSpeed[1] TSOcap[1]