[tulip-bug] AMDTek Tx Errors and Carrier Errors x2
bcarter@gumdrop.flyinganvil.org
bcarter@gumdrop.flyinganvil.org
Mon Nov 11 23:57:01 2002
> Many times there are multiple postings if I don't have time to track
> down the problem immediately, but same user doesn't have a motivation to
> post after the problem is solved.
>
> There is currently two outstanding issues with the ADMtek and Conexant
> chips where they report false Tx errors in full duplex mode. This does
> not impact communication, only the statistics.
I have a Linksys Tulip clone that uses the ADMtek chip and it reports
bogus statistics as well:
eth0 Link encap:Ethernet HWaddr 00:04:5A:6C:63:E4
inet addr:192.168.20.69 Bcast:192.168.20.79 Mask:255.255.255.240
UP BROADCAST NOTRAILERS RUNNING MTU:1500 Metric:1
RX packets:4769 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:7756 dropped:0 overruns:0 carrier:15511
collisions:0 txqueuelen:100
Interrupt:11 Base address:0xd800
This is not limited to full-duplex mode, though, as this card is plugged
into a dumb 10baseT repeater, so it's operating in half-duplex (or at
least it appears to be!) Everything _seems_ to be ok, although this
output is bothersome. Also, while the counts go up, the "carrier"
number seems to be consistently equal to (2 * errors - 1).
I remember seeing something similar to this before, when trying to use a
Winbond interface...
Although the message a couple weeks ago where you mentioned it as a
full-duplex you wrote:
> > MII PHY #1 transceiver registers:
> > 1100 786d 001d 2411 05e1 0021 0004 2001
>
> OK, the transceiver thinks it is in full duplex mode, but...
How do you get full duplex from this? Wouldn't that be 0041 instead of
0021 for link partner register?
...Now I remember. This problem appeared when using a 2.1 kernel
instead of a 2.0 kernel (the statistics were correct when booting 2.0).
Is it possible that some data structure change is responsible for this?
This was way back before 2.2 was released (although the system I'm
currently having the issue on is a 2.2.14 kernel.)
In the unlikely event that the diagnostic information should prove
helpful in tracking this down, I have attached output from the
diagnostic program.
sasquatch:~# ./tulip-diag -aaemmf
tulip-diag.c:v2.15 9/23/2002 Donald Becker (becker@scyld.com)
http://www.scyld.com/diag/index.html
Index #1: Found a ADMtek AL985 Centaur-P adapter at 0xd800.
ADMtek AL985 Centaur-P chip registers at 0xd800:
0x00: fff98000 ffffffff ffffffff 07f99800 07f99a00 fc664010 ff972113 ffffffff
0x40: fffe0000 fff597f8 00000000 fffe0000 00000000 78010705 68a00000 00000008
Extended registers:
0x80: 00664010 03fe7fff a04c0085 15aeffff 00008000 07f99a70 07f99900 ffe0f000
0xa0: f0000000 6c5a0400 ffffe463 00000000 40000000 00000000 00000000 00000000
0xc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
0xe0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 30000027
Comet duplex is reported in the MII status registers.
Transmit started, Receive started.
The Rx process state is 'Waiting for packets'.
The Tx process state is 'Idle'.
The transmit threshold is 128.
Comet MAC address registers 6c5a0400 ffffe463
Comet multicast filter 0000000040000000.
EEPROM 64 words, 6 address bits.
Ethernet MAC Station Address 00:04:5a:6c:63:e4.
Default connection type 'Autosense'.
PCI IDs Vendor 1317 Device 0985 Subsystem 1317 0570
PCI min_grant 64 max_latency 128.
CSR18 power-up setting 0xa04c****.
MII PHY found at address 1, status 0x786d.
MII PHY found at address 2, status 0x786d.
MII PHY found at address 3, status 0x786d.
MII PHY found at address 4, status 0x786d.
MII PHY #1 transceiver registers:
1100 786d 001d 2411 05e1 0021 0004 2001
0000 0000 0000 0000 0000 0000 0000 0000
1001 0000 2037 205f 0000 000f 7490 0000
8191 6946 2c58 1326 8911 0444 0230 0000.
MII PHY #2 transceiver registers:
1100 786d 001d 2411 05e1 0021 0004 2001
0000 0000 0000 0000 0000 0000 0000 0000
1001 0000 0037 205f 0000 000f 7490 0000
8191 6946 2c58 1326 8911 0444 0230 0000.
MII PHY #3 transceiver registers:
1100 786d 001d 2411 05e1 0021 0004 2001
0000 0000 0000 0000 0000 0000 0000 0000
1001 0000 0037 205f 0000 000f 7490 0000
8191 6946 2c58 1326 8911 0444 0230 0000.
MII PHY #4 transceiver registers:
1100 786d 001d 2411 05e1 0021 0004 2001
0000 0000 0000 0000 0000 0000 0000 0000
1001 0000 0037 205f 0000 000f 7490 0000
8191 6946 2c58 1326 8911 0444 0230 0000.
--
-Ben Carter
Human beings, who are almost unique in having the ability to learn from
the experience of others, are also remarkable for their apparent
disinclination to do so. - Douglas Adams, "Last Chance to See"