[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [SOLVED] linux-image-2.6-486 vs. linux-image -2.6-<x>



Hugo Vanwoerkom wrote:
Hugo Vanwoerkom wrote:
Andrew Sackville-West wrote:
On Thu, Mar 01, 2007 at 06:18:13AM -0600, Hugo Vanwoerkom wrote:
Andrew Sackville-West wrote:
On Wed, Feb 28, 2007 at 10:44:33AM -0600, Hugo Vanwoerkom wrote:
Hi,

Is there a place where the difference between these latest kernel binary images is documented, other than looking thru the config files or the descriptions of the packages.

When I install Etch on my machine from the daily built, it pulls in -486.

With that and grub I can boot into a USB disk.

But when I install linux-image-2.6-k7 (my processor) the boot will fail, because he cannot find the root device, meaning the initrd failed somehow.



The issue is the use of an USB harddrive.

yes, this is more than trivial at this point, I think.

Linux-image-4-486 has no problems when used on a USB disk partition on which Etch is installed with the daily-built d-i. I always boots right.

However...

When I install the same linux-image-4-486 on an older partition that now runs 2.6.20-ck1 and only refers to the USB disk from /etc/fstab, the boot will fail 50% of the time because he just doesn't wait long enough for the device to show up and changing mkinitrd.conf with DELAY=10 and running update-initramfs -u did not change anything.

Also when I use yaird instead there is never any problem exept for the poor behavior of yaird: see http://wiki.debian.org/USB-HD_Boot_Full_Debian?highlight=%28USB%29
and:
http://lists.debian.org/debian-user/2006/05/msg01557.html



I'm sorry I'm very confused as to what exactly you are trying to do
here. Are you using /boot from the hard-disk and then using / on USB
or what?

A

I have a PATA disk (80GB) in a USB enclosure:
http://www.newegg.com/Product/Product.asp?Item=N82E16817145657

The problem exists in booting a Debian stock kernel *either* from the USB disk itself (it shows up as /dev/sda and I have 6 partitions on it)
*or* from another HD with the USB disk referenced in /etc/fstab.

If you boot from a USB disk partition with a stock kernel he cannot find the root device.

When you boot with a stock kernel from another HD and have the USB disk referenced then he reports a bad superblock on the device.

But it is a mixed bag: I get those errors with a fullblown Sid system ( i.e. everything installed including X). I do *not* get those errors when I do a minimum Etch install to the USB disk, dist-upgrade to Sid and install another stock kernel.

Neither do I have problems with my own compiled 2.6.20-ck1 kernel + yaird as initrd. (But yaird has other problems: I don't know how to upgrade the system with that as initrd) *That* was George Hein's solution.

I would prefer to run stock kernels, but under all conditions they have to "work" with the USB disk, it seems.


But see this:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=401916


But the patch fits but gets a PANIC!
So do this:

In you favorite editor edit /usr/share/initramfs-tools/scripts/local

and under line 9 add:

       echo "Self-imposed wait of 15s..."
        /bin/sleep 15.0

save and run update-initramfs -u

Viola! Now you can run any Debian stock kernel on a USB disk.
It will show the message, wait 15secs (actually 10 would do) and merrily load any USB disk files.

Must say that not to many people passed that test ;-)

Hugo




















Reply to: