On Thu, 30 Mar 2006 05:34:09 -0800, yoginho wrote:
> Hi there,
>
> I've got a Tyan Thunder K8W motherboard with a Broadcom BCM5703C
> Gigabit ethernet card. My box is running Kubuntu 5.10 (Breezy Badger)
> with the tg3 driver:
>
> # uname -a
> Linux <hostname> 2.6.12-10-amd64-generic #1 Sat Mar 11 16:15:30 UTC
> 2006 x86_64 GNU/Linux
> # sudo ethtool -i eth0
> driver: tg3
> version: 3.31
> firmware-version:
> bus-info: 0000:02:09.0
>
> Whenever I transfer large amounts of data from my laptop to this
> machine via ssh, I get a
>
> 'Received disconnect from <hostname>: 2: Corrupted MAC on input.'
>
> usually fairly early during the transfer. I don't think this is a bug
> in OpenSSH (I'm using: OpenSSH_4.1p1 Debian-7ubuntu4.1, OpenSSL 0.9.7g
> 11 Apr 2005).
>
> I've found quite a bit of information on this error online but none of
> it is relevant to me, since neither do I have any linksys routers on
> the network, nor did switching off checksumming on the interface
> ('ethtool -K eth0 tx off rx off') do any good.
>
> The problem happens only when I transfer data over the local network,
> but not if I transfer data from home (slow, since via cable modem).
> This makes me think that the problem has something to do with high
> transfer speed.
>
> See below for the precise settings of the NIC. I have tried to lower
> the card's speed by using 'sudo ethtool -s eth0 speed 10', but this
> does not seem to have any effect. Anyway, permanently lowering the
> interface's speed does not seem very appealing to me.
>
> Any suggestions? Is this a known bug in the tg3 driver? I haven't come
> across it anywhere. Is it worth trying the outdated bcm5700 driver for
> Broadcom 57** cards, instead of tg3?
>
> Thanks for your help.
> Yogi
>
> # ifconfig eth0
> eth0 Link encap:Ethernet HWaddr 00:E0:81:34:38:90
> inet addr:131.111.111.245 Bcast:131.111.255.255
> Mask:255.255.255.0
Really? A real internet address and broadcast not on the same subnet?
> inet6 addr: fe80::2e0:81ff:fe34:3890/64 Scope:Link
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:157340 errors:0 dropped:0 overruns:0 frame:0
> TX packets:27600 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:78989270 (75.3 MiB) TX bytes:2088172 (1.9 MiB)
> Interrupt:24
>
> # sudo ethtool eth0
> Settings for eth0:
> Supported ports: [ MII ]
> Supported link modes: 10baseT/Half 10baseT/Full
> 100baseT/Half 100baseT/Full
> 1000baseT/Half 1000baseT/Full
> Supports auto-negotiation: Yes
> Advertised link modes: 10baseT/Half 10baseT/Full
> 100baseT/Half 100baseT/Full
> 1000baseT/Half 1000baseT/Full
> Advertised auto-negotiation: Yes
> Speed: 100Mb/s
> Duplex: Full
> Port: Twisted Pair
> PHYAD: 1
> Transceiver: internal
> Auto-negotiation: on
> Supports Wake-on: g
> Wake-on: d
> Current message level: 0x000000ff (255)
> Link detected: yes
>
Note: comment inline.
Where did you get the TPC/IP address you are using? Usually,
connections to the network are using a small router using NAT. What are
you connecting with?
Also, I'll bet if you check the kernel changelogs for kernels between
2.6.12 and 2.6.16 there will be quite a few matches for tg3. It remains in
flux.
I had a problem with a similar laptop connecting to a hotel's network in
Las Vegas. The problem was solved by checking out.
--
/* The Whole Year Inn. */
/* The hole you're in. */
http://us.imdb.com/title/tt0113627/quotes
Similar Threads
Linear Mode
