[Qemu-devel] root userid remapping? ... no UID / GID variables set in NFS root?

[Qemu-devel] root userid remapping? ... no UID / GID variables set in NFS root?

[Christopher Friedt]

Hi everyone,

I've just gotten qemu emulating an arm system on my laptop, which I'm 
hoping to use for cross-compilation. I don't have a disk image to use 
(is there any way to create a disk image from a directory tree??), so 
I'm using NFS instead.

The strange thing is, is that although I specify no_root_squash in my 
/etc/exports (and I've been using NFS for quite a long time successfully 
with other machines), only when I'm using Qemu does the user id of the 
'init' process [0] get remapped to some completely absurd value of 
4294967295.

Has anyone heard of this happening before?

Thanks in advance for any and all responses.

~/Chris

Re: [Qemu-devel] root userid remapping? ... no UID / GID variables set in NFS root?

[Mark Glines]

On Fri, 01 Jun 2007 00:06:43 +0200
Christopher Friedt <cfriedt@visible-assets.com> wrote:

> (is there any way to create a disk image from a directory tree??)

http://www.hohnstaedt.de/e2fsimage/

If you use this, you will get a raw e2fs filesystem image.  It will not
have a partition table, which means you will want to mount it directly
as /dev/hda (or /dev/sda), not /dev/hda1 (or /dev/sda1).


> The strange thing is, is that although I specify no_root_squash in my 
> /etc/exports (and I've been using NFS for quite a long time
> successfully with other machines), only when I'm using Qemu does the
> user id of the 'init' process [0] get remapped to some completely
> absurd value of 4294967295.

4294967295 is decimal for 0xffffffff, also known as -1.  Maybe some
kind of bug?

Mark

Re: [Qemu-devel] root userid remapping? ... no UID / GID variables set in NFS root?

[Rob Landley]

On Thursday 31 May 2007 6:06 pm, Christopher Friedt wrote:
> Hi everyone,
> 
> I've just gotten qemu emulating an arm system on my laptop, which I'm 
> hoping to use for cross-compilation. I don't have a disk image to use 
> (is there any way to create a disk image from a directory tree??), so 
> I'm using NFS instead.
> 
> The strange thing is, is that although I specify no_root_squash in my 
> /etc/exports (and I've been using NFS for quite a long time successfully 
> with other machines), only when I'm using Qemu does the user id of the 
> 'init' process [0] get remapped to some completely absurd value of 
> 4294967295.

You disabled CONFIG_UID16 in the EMBEDDED menu, so you don't have the legacy 
system calls to query 16 bit user ids, but your program is using the old ones 
rather than the new ones.  So the attempt to make said syscall is 
returning -EPERM, I.E. -1.

Nothing to do with qemu, this is your kernel .config.  I'm guessing you built 
a different kernel for qemu than for your host system?

Rob