probs with DFE-580 TX

Frank Sander FrankSanderDo at vodafone.de
Sat Apr 16 19:45:04 UTC 2005


Hi folks,

I was working a bit on the prob:

I have merged together the sundance driver source of the 2.6.11 kernel 
(last update brought me that peace) and the one delivered with the 
Network card.

IT WORKS :-)

now my question is:
Can some body with more experience in driver progging than me have a 
look at that?
How can we make sure (if my merging is ok) to bring that in one of the 
next updates?

thanks
Frank

Frank Sander wrote:

> Hi folks,
>
> after checking several options in the modprobe.conf and also the 
> mentioned acpi=off if have to give up.
> nothing changed the situation decribed before.
>
> what I do not understand:
> I get the fault message and it looks on the traffic analyser that I 
> get it before the traffic stops =-O !?!?
> than the traffic stops (I guess due to the transmitter reset) and than 
> it comes back.
> Unfortunately I can not proper analyse the traffic outside the 
> interface because the machine is ~1000km away from me. Anyway I have 
> the feeling, that the timeout comes together with a set of collisions 
> on the network. Can that?
>
> Does any body have an idea how this time out is created in the module?
>
> see you all
> Frank
> Mary Ellen Foster wrote:
>
>>On 4/15/05, Frank Sander <FrankSanderDo at vodafone.de> wrote:
>>  
>>
>>>kernel 2.6.10-1.770_FC3smp on a pentium 4 HT board with MEGAraid controler.
>>>DFE-580TX 4 port ethernet gives me every few seconds
>>>
>>>Apr 15 12:01:33 greystar01 kernel: NETDEV WATCHDOG: eth0: transmit timed out
>>>Apr 15 12:01:33 greystar01 kernel: eth0: Transmit timed out, TxStatus 00
>>>TxFrameId 02, resetting...
>>>
>>>and resets the port.
>>>The Card is transfering Data but with the reset I always get the
>>>transfer down for a few seconds.
>>>    
>>>
>>
>>That sounds like *exactly* the same problem I'm currently having with
>>my Broadcom Ethernet card in FC4test2
>>(https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=154512) -- also
>>a Pentium 4 HT processor. In test1, I was able to make it go away by
>>adding "acpi=off" to the command line in Grub; maybe that will work
>>for you with your card too.
>>
>>MEF
>>
>>  
>>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/fedora-test-list/attachments/20050416/1dea31b3/attachment.htm>


More information about the fedora-test-list mailing list