2.4.10-pre13: Panic mounting initrd

2.4.10-pre13: Panic mounting initrd

[Andreas Steinmetz]

Cut and paste via paper clipboard:


RAMDISK: Compressed image found at block 0
Freeing initrd memory: 1850k freed
read_super_block: can't find a reiserfs filesystem on (dev 01:00, block 64,
size 1024)
read_super_block: can't find a reiserfs filesystem on (dev 01:00, block 8, size
1024)
Kernel panic: VFS: Unable to mount root fs on 01:00


My initrd is ext2 formatted.
2.4.10-pre12 works, 2.4.10-pre13 compiled with same kernel config as pre12.


Andreas Steinmetz
D.O.M. Datenverarbeitung GmbH

Re: 2.4.10-pre13: Panic mounting initrd

[Andreas Steinmetz]

On 21-Sep-2001 Nick Ivanter wrote:
> Andreas,
> 
> Looks like you somehow haven't compiled ext2fs support into your new kernel. 
> Try to check that.
> 
If things would be that easy...

Snippets from a slight modification to mount_root() in fs/super.c

        for (fs_type = file_systems ; fs_type ; fs_type = fs_type->next) {
printk("VFS: Processing %s\n",fs_type->name);
                if (!(fs_type->fs_flags & FS_REQUIRES_DEV))
                        continue;
                if (!try_inc_mod_count(fs_type->owner))
                        continue;
printk("VFS: Trying %s\n",fs_type->name);


And now the results during boot (snippets again):

VFS: Processing bdev
VFS: Processing proc
VFS: Processing sockfs
VFS: Processing tmpfs
VFS: Processing pipefs
VFS: Processing ext2
VFS: Trying ext2
VFS: Processing msdos
VFS: Trying msdos
VFS: Processing vfat
VFS: Trying vfat
VFS: Processing iso9660
VFS: Trying iso9660
VFS: Processing nfs
VFS: Processing reiserfs
VFS: Trying reiserfs
Kernel Panic: ...

This clearly shows that there's definitely something going _very_ wrong.
Best bet is loading/decompression of initrd. As far as I'm following the list
there were initrd modifications discussed/done by Alexander Viro/Andrea
Arcangeli.


Andreas Steinmetz
D.O.M. Datenverarbeitung GmbH

Re: 2.4.10-pre13: Panic mounting initrd

[Alexander Viro]

On Fri, 21 Sep 2001, Andreas Steinmetz wrote:

> This clearly shows that there's definitely something going _very_ wrong.
> Best bet is loading/decompression of initrd. As far as I'm following the list
> there were initrd modifications discussed/done by Alexander Viro/Andrea
> Arcangeli.

Not between -pre12 and -pre13, actually.  OTOH, new rd.c code _is_ new
and VM modifications between -pre12 and -pre13 might affect it.

rd.c _is_ screwed up right now, so I'd recommend to be careful with
vanilla -pre12 or -pre13.  I'm merging Andrea's stuff for normal
ramdisks with, erm, getting initrd into sane state, so as soon as
it gets local testing I'll post the patch.