started getting repeated "bge0: PHY read timed out" messages

[tequnix at frogmi.net]

> 
> After running fine for a while, my new server running 6.2-RELEASE  
> with latest security patches as of last Thursday or Friday started  
> giving the message
> 
> bge0: PHY read timed out
> 
> and I found the following in the system log
> 
> Mar  6 02:59:33 server kernel: bge0: link state changed to DOWN
> Mar  6 02:59:36 server kernel: bge0: link state changed to UP
> Mar  6 03:00:27 server kernel: bge0: link state changed to DOWN
> Mar  6 03:00:29 server kernel: bge0: link state changed to UP
> Mar  6 05:09:27 server kernel: bge0: PHY read timed out
> Mar  6 05:09:27 server last message repeated 3 times
> Mar  6 05:09:27 server kernel: bge0: link state changed to DOWN
> Mar  6 05:09:29 server kernel: bge0: PHY read timed out
> Mar  6 05:09:50 server last message repeated 79 times
> Mar  6 05:09:50 server kernel: bge0: watchdog timeout -- resetting
> Mar  6 05:09:50 server kernel: bge0: PHY read timed out
> Mar  6 05:09:50 server last message repeated 4 times
> Mar  6 05:09:50 server kernel: bge0: RX CPU self-diagnostics failed!
> Mar  6 05:09:50 server kernel: bge0: flow-through queue init failed
> Mar  6 05:09:50 server kernel: bge0: initialization failure
> Mar  6 05:09:52 server kernel: bge0: PHY read timed out
> Mar  6 05:10:23 server last message repeated 140 times
> Mar  6 05:12:24 server last message repeated 616 times
> Mar  6 05:22:26 server last message repeated 2580 times
> Mar  6 05:32:28 server last message repeated 2460 times
> Mar  6 05:42:30 server last message repeated 2540 times
> Mar  6 05:52:32 server last message repeated 2572 times
> Mar  6 06:02:33 server last message repeated 2652 times
> Mar  6 06:12:34 server last message repeated 2532 times
> Mar  6 06:22:35 server last message repeated 2516 times
> Mar  6 06:32:37 server last message repeated 2452 times
> Mar  6 06:42:38 server last message repeated 2520 times
> Mar  6 06:52:39 server last message repeated 2544 times
> Mar  6 07:02:39 server last message repeated 2919 times
> Mar  6 07:12:41 server last message repeated 2521 times
> Mar  6 07:22:43 server last message repeated 2524 times
> Mar  6 07:32:44 server last message repeated 2440 times
> Mar  6 07:42:45 server last message repeated 2524 times
> Mar  6 07:52:48 server last message repeated 2528 times
> Mar  6 08:02:50 server last message repeated 2656 times
> Mar  6 08:12:51 server last message repeated 2552 times
> Mar  6 08:22:53 server last message repeated 2528 times
> Mar  6 08:32:55 server last message repeated 2436 times
> Mar  6 08:42:57 server last message repeated 2536 times
> Mar  6 08:52:58 server last message repeated 2544 times
> Mar  6 08:58:55 server last message repeated 1631 times
> 
> This appears to be a HW problem at first look.  But when the server  
> boots, it works fine for a while (hours, days??)
> 
> Here is the id in the boot message
> 
> Mar  6 09:01:21 server kernel: bge0: <Broadcom BCM5705 A3, ASIC rev.  
> 0x3003> mem 0xfeab0000-0xfeabffff irq 16 at device 14.0 on pci1
> Mar  6 09:01:21 server kernel: miibus0: <MII bus> on bge0
> Mar  6 09:01:21 server kernel: brgphy0: <BCM5705 10/100/1000baseTX  
> PHY> on miibus0
> Mar  6 09:01:21 server kernel: brgphy0:  10baseT, 10baseT-FDX,  
> 100baseTX, 100baseTX-FDX, 1000baseTX, 1000baseTX-FDX, auto
> Mar  6 09:01:21 server kernel: bge0: Ethernet address: 00:e0:81:61:e9:a0
> 
> Is this some sort of SW driver issue or is it a HW issue at first  
> glance?  I remember kind of reading about some BGE issues a while back.
> 
> Thanks
> Chad

sorry, no solution here, just a `me too`:

FreeBSD 6.2-RELEASE #0 i386
(in my case, it started right after upgrading from 6.1-RELEASE to
6.2-RELEASE)

[..]
Mar  4 20:01:39 lyekka kernel: sk0: link state changed to DOWN
Mar  4 20:01:45 lyekka kernel: sk0: link state changed to UP
Mar  4 20:02:18 lyekka kernel: sk0: link state changed to DOWN
Mar  4 20:02:20 lyekka kernel: sk0: link state changed to UP
Mar  4 20:02:35 lyekka kernel: sk0: link state changed to DOWN
Mar  4 20:02:39 lyekka kernel: sk0: link state changed to UP
Mar  4 21:50:35 lyekka kernel: sk0: link state changed to DOWN
Mar  4 21:50:38 lyekka kernel: sk0: link state changed to UP
Mar  4 21:50:52 lyekka kernel: sk0: link state changed to DOWN
Mar  4 21:50:54 lyekka kernel: sk0: link state changed to UP
Mar  5 00:45:00 lyekka kernel: sk0: link state changed to DOWN
Mar  5 05:28:56 lyekka kernel: sk0: phy write timeout
[..]

card is:
[..]
skc0: <Linksys EG1032 Gigabit Ethernet> port 0xa800-0xa8ff mem
0xef004000-0xef007fff irq 12 at device 7.0 on pci0
skc0: SK-9521 10/100/1000Base-T Adapter rev. (0x1)
sk0: <Marvell Semiconductor, Inc. Yukon> on skc0
[..]

after some time (it differs, normally after a few days) it is not
possible to send/receive data via this interface.
`ifconfig sk0 down ; ifconfig sk0 up` helps

i notice this UP/DOWN flap when i start/shutdown another machine with
direct connection to that one (nve card on the other side, connected
with kat7 crossover patch cable)

if replaced the (sk) nic, no difference, so i guess this is not a
hardware problem ?

regards, 
reinhard

-- 
Canada Bill Jones's Motto:
	It's morally wrong to allow suckers to keep their money.

Canada Bill Jones's Supplement:
	A Smith and Wesson beats four aces.