qla2310 hba at boot time

Bob Kryger bobk at panix.com
Fri Nov 16 15:51:34 UTC 2007


I'd like to see them, thanks..

That was my initial problem. I loaded up the firmware, but did not 
rebuild the initrd. Then I rebuilt my initrd and then pulled the new one 
apart to verify that the firmware was in there, and it is. But I still 
get the same behavior.

What is confusing me is when is the load being done?  I see the nash 
script in the initrd has the insmod command in it and the firmware is in 
the initrd, but the firmware is not loaded. I don't actually need the 
qla for the boot as I am booting off an ATA disk.  I see what appears to 
be a load of module in the dmesg with the error about not finding the 
firmware, is this the message from the initrd sequence or some other 
part of the boot up?

So after working on this all day yesterday....

1. Is the output from the initrd loading and therefor the nash init 
script in the dmesg?
2. is the initrd the only place that would load this module, or is there 
some script after the initrd hands off to init that also loads such 
modules? Where might that be? I cannot find such a script.
3. what is the best way to go about traking this down - hack my initrd, 
or mkinitrd?
4. did i miss a step on regenerating my initrd?
5. I can certainly add a rc.local script or the like, but I'd rather 
this worked as intended, and I've rather been enjoying digging into 
parts of linux that I don't normally get to do. This is quite a change 
from adding users, and fixing permissions and such...

Thanks for your help
Bob

Mark Haney wrote:
> Andrew Parker wrote:
>> On Nov 15, 2007 2:33 PM, Bob Kryger <bobk at panix.com> wrote:
>>  
>>> So, I have an F7 system w/ a QLA2310 HBA. Everything works OK, when 
>>> done
>>> by hand.
>>>
>>> But when I reboot, the kernel module is not getting loaded correctly,
>>> the firmware not loaded, then the md not assembled and not mounted.
>>>
>>> So I unload the module, reload the module (which loads the firmware),
>>> assemble the md0, and mount by hand and it works.
>>>
>>> Questions:
>>>   Why is the module load at boot time not loading the firmware?
>>>   Where is that module load done in the startup sequence?
>>>   How do I approach resolution correctly, avoiding hacking the startup?
>>>     
>>
>> did you rebuild your initrd?  i seem to remember getting some
>> wierdness with our 2312 cards (on RHEL, not Fedora) like that
>>
>>   
> Yeah I had the exact same problem last week and rebuilding initrd 
> worked perfectly.  I have the instructions that someone posted (and 
> whoever it was thank you!) I'll be glad to send them to you.
>
>




More information about the fedora-list mailing list