Hello, Maks and others.
before jumping to strange conlcusions.
this looks very much like a dup.
did you set MODULES=dep *and* regenerated the initramfs with
update-initramfs -u -k <kernelversion>
*and* tried to boot that.
You're right, Maks, i_DID a mistake !
I'm not even able to describe what I did, but when I looked-at my
'MODULES=dep' - see 3) below - initrd I realized how ugly it was ; it
really contained evil things ! Strangely enough, my 'customized-dep' -
see 4) below - initrd version was completely fine !!!
If I didn't have made this mistake, I would have saved a few hours
guessing which few modules I had to add to this base.
Now - a few dozens minutes ago - I've rebuilt my 4 different versions of
initrd for linux-image-2.6.25-2-amd64 (ver 2.6.25-5) :
1) The default one (<=> Simply with MODULES=most)
2) A custom gently stripped-down version of the previous one
3) The 'dep' one (<=> Simply with MODULES=dep)
4) A (now useless) custom slighlty enhanced version of the previous one
And LILO knows them all very well ;-)
I've tried them all in order:
1) and 2) lead to an early kernel panic
3) and 4) are all fine wrt all one can expect from an initrd.
Now, that matter boils down to tree cases:
1) Linux has a problem with large initrd files.
2) LILO has a problem with large initrd files.
3) BOTH have a problem with large initrd files.
I know a snake who bites its tail ...
And now let's see my initrd sizes:
1) 7275892 bytes
2) 5738810 bytes
3) 3485038 bytes
4) 3686321 bytes
Note: vmlinuz is 1727488 bytes
But I can boot linux-image-2.6.24-1-amd64 (ver 2.6.24-7) without any
problem and using an initrd of 7090895 bytes, and vmlinuz is 1668248 bytes.
So, why had I an early kernel panic with version 2) (see above) of my
initrd for linux-image-2.6.25-2-amd64 (ver 2.6.25-5) which is smaller ?
In hope my report will prove useful.
I'm looking forward to hearing from you.
Sincerely,
Valentin QUEQUET
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]