<!-- SBZ_reproduce  -->
install SL 9.1 + the YOU update kernel (I didn't test it with the delivered
default kernel).

reassigning to the new maintainer

*** This bug has been marked as a duplicate of 58248 ***

Karl, can you reproduce the bug?

Still the same with STABLE from end of June.  I'll test it again on pre1 next week.

I have just fixed eject to use SG_IO (bug #113861), could you please test if the
eject now works?

Thanks for notification--I'll test it on Thursday.

any results?

Sorry for the delay.

'modprobe imm' (for loading the driver)
'mount /dev/sda1 /mnt

Then 'eject /dev/sda1' or 'eject /mnt' unmounts the disk.  If you call 'eject
/dev/sda1' a second time, it ejects the medium.

At the moment, the drive is connected to K21.suse.de (I can send you the
password on request).

Thanks for the offer, but for effective remote debugging it would have to be
able to load the medium without manual intervention :) Just please try with -v,
strace output could be interesting too.

Lada, could we obtain a SCSI ZIP for the testing?

;)

Oops, I forgot to mention that this time I run the test with a parport drive ;-(
 @home I have a SCSI drive...  I'm not sure how long it might take to update my
system at home from 9.3 to 10.0.  Sorry.

ok, let me then know how it works

tux:~ # strace -o e.log eject -v /dev/sda1
eject: device name is `/dev/sda1'
eject: expanded name is `/dev/sda1'
eject: `/dev/sda1' is mounted at `/mnt'
eject: unmounting `/dev/sda1'
eject: `/dev/sda1' is a multipartition device
eject: trying to eject `/dev/sda1' using CD-ROM eject command
eject: CD-ROM eject command succeeded

But it does not actually eject the medium.  The second time the following happens:

tux:~ # strace -o e2.log eject -v /dev/sda1
eject: device name is `/dev/sda1'
eject: expanded name is `/dev/sda1'
eject: `/dev/sda1' is not mounted
eject: `/dev/sda1' is not a mount point
eject: `/dev/sda1' is a multipartition device
eject: trying to eject `/dev/sda1' using CD-ROM eject command
eject: CD-ROM eject command succeeded

I'll attache e.og and e2.log

Created attachment 55260 [details]
e 1

Created attachment 55261 [details]
e 2

I have updated the eject in 10.1, does it still happen there?

any success with the 10.1?

assuming the bug is fixed in progress