(cc'ing linux-kernel as well since I suspect another subsystem may be
involved).

I noticed something else in addition to the problem below: I also
can no longer mount /dev/hdc1 since mount tells me the device
is busy. It looks like something scans the various block devices during
boot and forgets to release something.

Wichert.

Previously Wichert Akkerman wrote:

I am trying to setup RAID 1 array on two SCSI disks on a 2.6.0-test9
kernel but I am hitting the following:

vortex:~# mkraid /dev/md0
handling MD device /dev/md0
analyzing super-block
disk 0: /dev/hda2, 35077927kB, raid superblock at 35077824kB
disk 1: /dev/hdc2, 35077927kB, raid superblock at 35077824kB
md: bind<hda2>
md: could not lock hdc2.
md: error, md_import_device() returned -16
mkraid: aborted.
(In addition to the above messages, see the syslog and /proc/mdstat as well
 for potential clues.)

-16 is EBUSY, which is unepexted since nothing else is using that disk.
At this stage /proc/mdstat contains:

Personalities : [raid1] 
md0 : inactive hda2[0]
      35077824 blocks
unused devices: <none>

-- 
Wichert Akkerman [off-list ref]    It is simple to make things.
http://www.wiggy.net/                   It is hard to make things simple.