Re: Creating working qemu image

[Allan McRae]

On 26/01/10 01:55, Samuel Thibault wrote:
> olafBuddenhagen@gmx.net, le Mon 25 Jan 2010 12:07:48 +0100, a écrit :
>    
> > On Sun, Jan 17, 2010 at 01:01:09AM +1000, Allan McRae wrote:
> >
> >      
> > > start /hurd/ext2fs.static: ext2fs: device:hd0s1: panic:
> > > get_hypermetadata: disk size (1069254144 bytes) too small; superblock
> > > says we need 1073709056
> > >        
> > I don't think I have seen something like this before; but from the sound
> > of it, I'd say something went wrong with the mkfs? Perhaps something
> > wrong with the loopback setup? I'm not convinced this is really a Hurd
> > problem...
> >      
> Looks like mkfs promised more than the partition could actually contain.
> How did you run it?
>    


I created it using this:

     qemu-img create -f raw hurd.img 1GB
     losetup /dev/loop0 hurd.img
     fdisk /dev/loop0  (create one partition)
     losetup -d /dev/loop0
     losetup -o 32256 /dev/loop0 hurd.img
     mkfs.ext2 -b 4096 -I 128 -o hurd -F /dev/loop0

> To make it simple, I'd advise running mkfs again, giving it 261048
> (i.e. 1069254144/4096) as number of blocs.
>    


Good idea.  I will try that later.  At the moment I am just using the
Debian Hurd image and deleting all the files...

Thanks,
Allan