Skip to contentSkip to conversation

VIA Rhine ethernet driver bug (reprise)

DISCUSSION Udo van den Heuvel 2 participants 2 messages Aug 7, 2005 – Aug 11, 2005

lore
Share & tools
lore
#0 lore raw

Hello,

In january of this year I mentioned a problem with the Linux kernel driver for VIA Rhine ethernet chips. (see http://lkml.org/lkml/2005/1/15/47)
In the mean time this bug was reproduced quite a number of times on my Fedora Core 3 (then)/4 (now) box (a VIA CL6000).

An alternative driver by VIA was used (with kernel 2.6.1x,
http://www.viaarena.com/downloads/Source/rhinefet.tgz); this VIA driver did not have the Oversized ethernet frame bug but consumes quite a lot of CPU when transfering a steady stream of a few 1000 KB/S, so it is no good solution.

Since the ethernet connection goes down due to this bug I think I can mark the bug as critical. The bug is still present in 2.6.12. The maintainer of the driver cannot fix the bug. Who can?

Please post here and/or email me with your thoughts about a solution, ways of fixing the bug. Would a small reward help?

Kind regards,
Udo

#1 lore raw
Quoted 13 lines · Udo van den Heuvel

Hello,

In january of this year I mentioned a problem with the Linux kernel driver for VIA Rhine ethernet chips. (see http://lkml.org/lkml/2005/1/15/47)
In the mean time this bug was reproduced quite a number of times on my Fedora Core 3 (then)/4 (now) box (a VIA CL6000).

An alternative driver by VIA was used (with kernel 2.6.1x,
http://www.viaarena.com/downloads/Source/rhinefet.tgz); this VIA driver did not have the Oversized ethernet frame bug but consumes quite a lot of CPU when transfering a steady stream of a few 1000 KB/S, so it is no good solution.

Quoted 3 lines

Since the ethernet connection goes down due to this bug I think I can mark the bug as critical. The bug is still present in 2.6.12. The maintainer of the driver cannot fix the bug. Who can?

at http://lkml.org/lkml/2005/1/15/47:

Quoted 1 line

I see messages like:

Jan 13 19:35:46 epia kernel: eth1: Oversized Ethernet frame spanned multiple buffers, entry 0x4 length 0 status 00000600!
Jan 13 19:35:46 epia kernel: eth1: Oversized Ethernet frame ccf80040 vs ccf80040.
Jan 13 19:35:46 epia kernel: eth1: Oversized Ethernet frame spanned multiple buffers, entry 0x5 length 0 status 00000400!
...
Jan 13 19:35:46 epia kernel: eth1: Oversized Ethernet frame spanned multiple buffers, entry 0x3 length 0 status 00000581!
Jan 13 19:35:46 epia kernel: eth1: Oversized Ethernet frame ccf80030 vs ccf80030.

every 3 or 4 days or so when I really use the card. (please notice all 16 entries are used and the length is 0)

Quoted 1 line

While googlin' I saw that this is an old bug which was not fixed for years.

I am in contact with the maintainer and currently am trying to see what effect a smaller mtu and/or older driver versions might have.

Quoted 1 line

Because of the impact of this nasty bug (many users have these chips in

their hardware) I would like to ask if others could have a look into this problem as well. Please email me your experiences (2.6.x kernel? Same bug or no problem? Fixes? Etc).

What is "ccf80030 vs ccf80030"? Those numbers are the same... Maybe you need to add more debugging prints in that part of code.

But anyway:

Since it happens less than once a day, why not just add a code to reset the NIC completely in this case, like it is typically done in tx_timeout handlers of many NICs, and forget about it? --
vda