gvfs access inexistent fd0 drive

Could you please post the output of:

devkit-disks --dump
gvfs-mount -l

Thanks

As you may see in dmesg, there is an fd0 controller detected but there is no physical drive.
Following output comes from "modprobe floppy"

[ 2.031611] Floppy drive(s): fd0 is 1.44M
[ 2.047060] FDC 0 is a National Semiconductor PC87306

# ps ax | egrep fd0 | egrep -v egrep
 3772 ? D 0:00 devkit-disks-daemon: polling /dev/sr0 /dev/fd0

There is no output from gvfs-mount -l

Also when using "modprobe floppy floppy=debug" I just have following trace:

[ 2534.816115] floppy0: Setting flag 0x1
[ 2534.816250] Floppy drive(s): fd0 is 1.44M
[ 2534.832040] floppy0: reschedule timeout lock fdc
[ 2534.833468] reset interrupt: dtime=558708
[ 2534.833480] floppy0: reschedule timeout do wakeup
[ 2534.833793] FDC 0 is a National Semiconductor PC87306
[ 2534.833813] floppy0: reschedule timeout lock fdc
[ 2534.834699] reset interrupt: dtime=558708
[ 2534.834711] floppy0: reschedule timeout do wakeup

But afterwards, dmesg is permanently filled in by following output just saturating the ring buffer

[ 2541.096925] floppy0: checking disk change line for drive 0
[ 2541.096933] floppy0: jiffies=560274
[ 2541.096943] floppy0: disk change line=0
[ 2541.096949] floppy0: flags=2
[ 2541.096967] queue fd request dtime=1555
[ 2541.096973] floppy0: reschedule timeout floppy start
[ 2541.096982] floppy0: setting NEWCHANGE in floppy_start
[ 2541.116217] floppy0: calling disk change from floppy_ready
[ 2541.116226] floppy0: checking disk change line for drive 0
[ 2541.116233] floppy0: jiffies=560279
[ 2541.116243] floppy0: disk change line=0
[ 2541.116249] floppy0: flags=6
[ 2541.116293] floppy0: calling disk change from seek
[ 2541.116300] floppy0: checking disk change line for drive 0
[ 2541.116306] floppy0: jiffies=560279
[ 2541.116316] floppy0: disk change line=0
[ 2541.116322] floppy0: flags=2
[ 2541.116332] recalibrate floppy: dtime=1560
[ 2541.465651] recal interrupt: dtime=1647
[ 2541.465659] recal interrupt need 1 recal: dtime=1647
[ 2541.465672] floppy0: reschedule timeout redo fd request

I am also experiencing repeated attempts to access /dev/fd0. Since my floppy drive actually exists, but there is no drive, I am going crazy with its constant clicking. Running 'sudo modprobe -r floppy' gets rid of the clicking, but the drive light stays on, which makes me wonder how clean the unloading of floppy was.

I'm providing the same information as the original reporter for comparison:

$ apt-cache policy gvfs
gvfs:
  Installed: 1.3git20090512-0ubuntu2
  Candidate: 1.3git20090512-0ubuntu2
  Version table:
 *** 1.3git20090512-0ubuntu2 0
        500 http://archive.ubuntu.com karmic/main Packages
        100 /var/lib/dpkg/status
$ dmesg | tail
[ 348.836058] end_request: I/O error, dev fd0, sector 0
[ 350.836031] end_request: I/O error, dev fd0, sector 0
[ 352.836049] end_request: I/O error, dev fd0, sector 0
[ 354.836039] end_request: I/O error, dev fd0, sector 0
[ 356.836081] end_request: I/O error, dev fd0, sector 0
[ 358.836056] end_request: I/O error, dev fd0, sector 0
[ 360.836031] end_request: I/O error, dev fd0, sector 0
[ 362.836031] end_request: I/O error, dev fd0, sector 0
[ 364.836038] end_request: I/O error, dev fd0, sector 0
[ 366.836060] end_request: I/O error, dev fd0, sector 0
$ devkit-disks --dump > devkit-disks # see attached file
$ gvfs-mount -l
Drive(0): (null)
  Type: GProxyDrive (GProxyVolumeMonitorGdu)
Drive(1): TEAC DVD+RW DV-W58E
  Type: GProxyDrive (GProxyVolumeMonitorGdu)
$ sudo modprobe -r floppy
$ sudo modprobe floppy floppy=debug
$ dmesg | tail -n 40 # after waiting a bit
[ 632.836045] floppy0: calling disk change from seek
[ 632.836049] floppy0: checking disk change line for drive 0
[ 632.836052] floppy0: jiffies=83209
[ 632.836063] floppy0: disk change line=80
[ 632.836066] floppy0: flags=2
[ 632.836078] floppy0: reschedule timeout request done 0
[ 632.836086] end_request: I/O error, dev fd0, sector 0
[ 634.817327] floppy0: reschedule timeout lock fdc
[ 634.817388] floppy0: reschedule timeout redo fd request
[ 634.817395] floppy0: checking disk change line for drive 0
[ 634.817399] floppy0: jiffies=83704
[ 634.817430] floppy0: disk change line=80
[ 634.817433] floppy0: flags=2
[ 634.817455] queue fd request dtime=34421
[ 634.817459] floppy0: reschedule timeout floppy start
[ 634.817464] floppy0: setting NEWCHANGE in floppy_start
[ 634.817468] floppy0: calling disk change from floppy_ready
[ 634.817472] floppy0: checking disk change line for drive 0
[ 634.817476] floppy0: jiffies=83704
[ 634.817486] floppy0: disk change line=80
[ 634.817489] floppy0: flags=6
[ 634.817501] floppy0: calling disk change from seek
[ 634.817599] floppy0: checking whether disk is write protected
[ 634.817602] floppy0: wp=40
[ 634.817699] seek command: dtime=34421
[ 634.822409] seek interrupt: dtime=34422
[ 634.822415] floppy0: clearing NEWCHANGE flag because of effective seek
[ 634.822419] floppy0: jiffies=83705
[ 634.840039] floppy0: calling disk change from floppy_ready
[ 634.840048] floppy0: checking disk change line for drive 0
[ 634.840052] floppy0: jiffies=83710
[ 634.840064] floppy0: disk change line=80
[ 634.840068] floppy0: flags=0
[ 634.840081] floppy0: calling disk change from...

Jacob - your issue is a separate one with devicekit-disks where it polls floppy drives. I'm not sure if a bug report for that exists already, but I mentioned it to davidz a few weeks ago on IRC, and he's already aware of your problem and it's likely to be fixed during this cycle.

nmn - your issue looks more like a kernel bug rather than a userspace bug, as it is enumerating a floppy drive when one doesn't exist.

I agree.

A work around is to blacklist the non-existent floppy:
add to /etc/modprobe.d/blacklist.conf the lines

 # floppy driver since I don't have one
 blacklist floppy

and then run update-initramfs.

Thanks.