This issue has been proposed when we are only considering blocker
issues in the current Red Hat Enterprise Linux release.

** If you would still like this issue considered for the current
release, ask your support representative to file as a blocker on
your behalf. Otherwise ask that it be considered for the next
Red Hat Enterprise Linux release. **

Well, that works for me fine, I can't reproduce it with the
current stack:

[root@test tmp]# virsh attach-device test6.1 disk-block.xml
Device attached successfully

[root@test tmp]# getenforce
Enforcing
[root@test tmp]# rpm -q libvirt qemu-kvm qemu-img
libvirt-0.8.1-20.el6.x86_64
qemu-kvm-0.12.1.2-2.104.el6.x86_64
qemu-img-0.12.1.2-2.104.el6.x86_64
[root@test tmp]# ls -l /dev/sdb1
brw-rw----. 1 qemu qemu 8, 17 Jul 31 02:45 /dev/sdb1
[root@test tmp]# virsh detach-device test6.1 disk-block.xml
Device detached successfully

[root@test tmp]# ls -l /dev/sdb1
brw-rw----. 1 root disk 8, 17 Jul 31 03:28 /dev/sdb1
[root@test tmp]# cat disk-block.xml
<disk type='block' device='disk'>
  <source dev='/dev/sdb1'/>
  <target dev='vdb' bus='virtio'/>
</disk>


[root@test tmp]# 


In the guest I saw in dmesg the virtio-pci logs ending up with
 vbb: unknown partition table.

I also tried with an /dev/sr0 device read-only that worked fine too
and the attachemnt of a file based device also worked fine.

Could you double check with the recent stack as this seems to work
fine for me with current packages.
Also what kind of guest are you using ?
And did you get any error on the qemu logs ?

Daniel

I logged in the box yesterday and could reproduce the issue
it's very strange indeed, nothing seems to be wrong ... except that
specific command doesn't work, I will try to debug the issue further if you can keep the access to that machine and setup for me.
One question though, the guest is running for a week now, did you update any
of the libvirt or kvm packages on that box since then ?

Daniel

> > And did you get any error on the qemu logs ?
> 
> /var/log/libvirt/qemu/rhel6beta2.log
> ----------
> qemu: could not open disk image /dev/sda3: Operation not permitted

What kernel version do you have, and are any of the 'cgroup' filesystems shown in /proc/mounts ?

I expect this is a duplicate of this kernel bug.

https://bugzilla.redhat.com/show_bug.cgi?id=589662

I logged on the test machine, stopped cgconfig, 
[root@ibm-hs21-1 tmp]# /etc/init.d/cgconfig stop
Stopping cgconfig service:                                 [  OK  ]
[root@ibm-hs21-1 tmp]# /etc/init.d/libvirtd restart
Stopping libvirtd daemon:                                  [  OK  ]
Starting libvirtd daemon:                                  [  OK  ]
[root@ibm-hs21-1 tmp]# 
[root@ibm-hs21-1 tmp]# virsh attach-device rhel6beta2 disk-block.xml
Device attached successfully

[root@ibm-hs21-1 tmp]# 


so that could be the problem, yes

[root@ibm-hs21-1 tmp]# ls /cgroup/
[root@ibm-hs21-1 tmp]# 

Daniel

*** This bug has been marked as a duplicate of bug 589662 ***