Actually, I had seen this behavior before, seemingly sporadically, but
now that I look back, and had some time to investigate, it made sense.
I have various machines with hardware for several of the kernel modules
that we package up in the kernel-modules rpm's.
The reason this was sporatic, was because I would often be in the
correct kernel, and the scripts from the rpm's, would correctly
configure the kernel module. But occasionally I wouldn't be, and then
I'd get this mysterious problem that François would see. I'd uninstall
then re-install the rpm, it would work, and I'd usually get pulled off
to work on something else before I could think about it.
Since nobody else had the problem, it never come up high enough in my
priorities.
I've re-done most of the kernel-module rpm's for the next kernel errata
comming out. I patterned the post install script after the
openafs-kernel module, since that is, by far, the most used
kernel-module rpm. In my tests, it's worked so far.
Troy
Donald E Tripp wrote:
> Very strange, since dmesg should show the card even if it doesn't have the correct driver. Is this happening one a single
> machine, or multiple machines? That may give you a better idea of where to start, and how to narrow things down.
>
>
> - Don
>
> ----- Original Message -----
> From: Francois Colonna <[log in to unmask]>
> Date: Thursday, October 5, 2006 9:32 pm
> Subject: no network with kernel 42.0.2. please answer
> To: [log in to unmask]
>
>> Hello
>>
>>
>> with kernel 42.0.2.ELsmp
>> it is impossible for me to make the network card (Realtek) be
>> recognized
>> dmesg | grep eth0
>>
>> gives an empty answer
>>
>> while with the 2.6.9-34.0.2.ELsmp kernel it gave :
>>
>> eth0: Identified chip type is 'RTL8168B/8111B'.
>> eth0: r10001.0, the Linux device driver for Realtek Ethernet
>> Controllers at
>> 0xc800, 00:17:31:25:6b:37, IRQ 169
>> eth0: Auto-negotiation Enabled.
>> eth0: 100Mbps Full-duplex operation.
>> eth0: no IPv6 routers present
>>
>> How to make things work ?
>>
>>
>> Thanks
>> François Colonna
>>
--
__________________________________________________
Troy Dawson [log in to unmask] (630)840-6468
Fermilab ComputingDivision/CSS CSI Group
__________________________________________________
|