2.6 kernel (was Hardware detection hint)

Declan Moriarty declan.moriarty at gmail.com
Fri Apr 15 08:52:50 PDT 2005


All this from 5 months ago.....

I was having problems with cursors on tty2 --> and the 2.6 kernel and
then was looking for explanations of strange behaviour. Now I think I
have it.
> 
> > echo -e '\033[?25h'  did nothing.
> 
> Strange. Let's then (just in case) check things that differ in our setups,
> even though I think they are irrelevant.
> 
> 1) I execute the "setfont" command during boot as a part of the "console"
> bootscript. Does sticking a "setfont" without parameters into some bootscript
> help in your case?
What appears to be happening on the 2.6.20 kernel is this: The
directory /etc/rc.d/rcsysinit.d has links to init.d scripts.
S40mountfs now seems to run OK( I did have trouble with some
partitions not mounting until I changed permissions.) I am in windows
currently - I'll change back and investigate the mountfs script
latter.
With the 2.6 kernel,  everything after that is not run. In my version
of the scripts I that includes the loadkeys script. Simply typing 
"setfont" restores the cursor, because this is the first time it has
been run since booting.
I haven't a list of what is up there but localnet(setting up loopback)
is among them, as is setting the clock and some check on the file
systems There is a plaintive cry in sys.log asking "How do I bring up
network interfaces in this distro?" in every 2.6 bootup but typing
/etc/rc.d/init.d/localnet start (BLFS 5) brings up interfaces all
lovely. I did these checks

dmesg |grep loopback  from the localnet script.
dmesg |grep clock from the setclock script
dmesg |grep forcefsck from S50cleanfs or something like that

and found nothing - nada

That's what is happening.  Why, I don't know. If anyone read this far,
it's generally not a good idea that scripts can exit quietly without
noticeable errors in a bootup, is it? What happened to the red writing
and threatening noises we come to expect??

Apart from that, life is normal. Still in search of a Holy Grail here
(pppoe access via usb).
usbutils will not compile :-(. It finds unresolved hotplug symbols. 
/proc/bus/usb won't mount because of hotpug symbols either. I grabbed
hotplug, but no dice. 2.4 kernel compiles threw up nasties and I
decided to stay with 2.6 if possible. pppoe seems promising and did
compile.

Am I totally out of date with usbutils-0.11, hotplug-20040329?
-- 
     All the Best

     Declan.



More information about the blfs-support mailing list