msk nic and watchdog timeouts

[Leonidas Tsampros]

Hello,

I ran a machine with FreeBSD 7.0-PRERELEASE (i386) (checked out and
built on 12 January). My mainboard has 2 onboard NICs, which are
identified correctly:

Feb 24 18:54:35 bifteki kernel: msk0: <Marvell Technology Group Ltd. Yukon EC Id 0xb6 Rev 0x01> on mskc0
Feb 24 18:54:35 bifteki kernel: sk0: <Marvell Semiconductor, Inc. Yukon> on skc0

Output from pciconf -lv:

mskc0 at pci0:2:0:0:       class=0x020000 card=0x100a15bd chip=0x436211ab rev=0x19 hdr=0x00
    vendor     = 'Marvell Semiconductor (Was: Galileo Technology Ltd)'
    device     = '88E8053 Marvell Yukon 88E8053 PCI-E Gigabit Ethernet Controller'
    class      = network
    subclass   = ethernet
skc0 at pci0:3:10:0:       class=0x020000 card=0x100a15bd chip=0x432011ab rev=0x13 hdr=0x00
    vendor     = 'Marvell Semiconductor (Was: Galileo Technology Ltd)'
    device     = 'Yukon 88E8001/8003/8010 PCI Gigabit Ethernet Controller (Copper)'
    class      = network
    subclass   = ethernet

After random intervals of time and usage the msk0 driver fails with
the followings messages:

Feb 24 18:27:21 bifteki kernel: msk0: watchdog timeout
Feb 24 18:27:21 bifteki kernel: msk0: link state changed to DOWN
Feb 24 18:27:23 bifteki kernel: msk0: link state changed to UP
Feb 24 18:28:02 bifteki kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Feb 24 18:28:15 bifteki kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Feb 24 18:28:20 bifteki kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Feb 24 18:28:28 bifteki kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering

and after this point the nic becomes unusable.

Has anyone noticed this happening? Is there any way to help debugging
this issue ? I thought of raising dev.mskc.%d.process_limit, but I
don't know if that's related.

Thanks,