From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.7) Gecko/20050416 Fedora/1.0.3-1.3.1 Firefox/1.0.3 Description of problem: Plugging and unplugging a USB key in GNOME seems to cause untold damage in the latest released Fedora Core 3 kernel. (With nvidia binary drivers the system locks up as soon as X is quit) Without any binary drivers (tested on an Duron with the nvidia driver *not loaded* and on a K6-II with an ATI Rage card) the following appears on shutdown: Kernel panic - not syncing: drivers/block/cfq-iosched.c:1065: spin_is_locked on unititialized spinlock c5551c1c. (Not tainted) (that was the Duron, the K6 said nearly the same but the address was c54aa81c) Older kernels (e.g. kernel-2.6.10-1.770_FC3) do not seem to have this problem. Version-Release number of selected component (if applicable): kernel-2.6.11-1.14_FC3 How reproducible: Always Steps to Reproduce: 1. Start the machine. 2. Log in to GNOME. 3. Plug USB key in and wait for USB key icon to appear on the desktop. 4. Press the right mouse button over the icon and go to Unmount. 5. Wait for icon to disappear. 6. Go to Actions -> Log Out. Select "Restart the computer" and click OK. Actual Results: A few services are shutdown (the last being cups) and then the following is printed: Kernel panic - not syncing: drivers/block/cfq-iosched.c:1065: spin_is_locked on unititialized spinlock c5551c1c. (Not tainted) Expected Results: System should have shut down cleanly. Additional info: Quite possibly related to / a dup of bug #154764 (although I'd like to avoid the nvidia binary driver controvesy) or bug #155381 (the line number is different but then so is the kernel version).
Bug #154690 is also a good dup candidate.
I forgot one key step: 4.5. Physically remove the key from the machine. Without this step the problem will not manifest itself. I have tested this on a newer kernel-2.6.11-1.1251_FC4 (PS this kernel seems to have some extra selinux stuff that causes trouble on FC3 in enforcing mode due to xyz producing denied errors) and this is the output: Kernel panic - not syncing: drivers/block/cfq-iosched.c:1070: spin_is_locked on unititialized spinlock c55b6348. (Not tainted) [<c01201db>] panic+0x3b/0x1b0 [<c028ef37>] cfq_exit_io_context+0x127/0x3b0 [<c0287a2f>] exit_io_context+0x3f/0x50 [<c0123d72>] do_exit+0x42/0x4d0 [<c01242db>] do_group_exit+0x9b/0x1b0 [<c01039e9>] syscall_call+0x7/0xb
Can you try the test kernel from http://people.redhat.com/davej/kernels/Fedora/FC3/RPMS.kernel/ and see if that fixes it ?
I'm seeing the same problem with an external USB 2.0 Lacie hard-drive. Steps to reproduce: 1. Plug the drive or boot the machine with it already plugged in 2. mount 3. unmount 4. turn off the drive 5. reboot 6. (kernel panic)
I tried with 2.6.11-1.19_FC3, looks fixed. :)
(sorry for the sloth like response on this) Tested with two different keys on 2.6.11-1.19_FC3 and the problem errors on shutdown no longer appear. The smug people who have been locking up machines with their USB keys have been thwarted and davej saves the day again. Thanks for the fast turn around. Let's see, how shall we resolve this...
*** Bug 156050 has been marked as a duplicate of this bug. ***
reopening until this gets pushed out as an update.
*** Bug 156593 has been marked as a duplicate of this bug. ***
Dave Jone's Kernel works for me and for many others. Where's the problem to release a fixed kernel? USB-Sticks/devices are used very heavily. This bug damages fedora's reputation a lot!?
2.6.11-1.21_FC3 fixees the problem I was seeing in bug #156593.
*** Bug 157016 has been marked as a duplicate of this bug. ***
*** Bug 156750 has been marked as a duplicate of this bug. ***
*** Bug 155961 has been marked as a duplicate of this bug. ***
*** Bug 157426 has been marked as a duplicate of this bug. ***
*** Bug 158326 has been marked as a duplicate of this bug. ***
I suggest that 'kernel panic' be put in the subject somewhere to avoid dups.
(In reply to comment #17) > I suggest that 'kernel panic' be put in the subject somewhere to avoid dups. I'd suggest releasing a working kernel. I don't have this problem anymore with Dave Jones's kernel (see comment 3).
I'm normally loathe to change subjects because it causes unecessary spam to the people cc'd, makes subjects so long they don't fit neatly in the bug lists and leads to subject tug of wars (plus I did include a bit of the error for people to search on AND if you do a search on body text then the words kernel panic will show up PLUS if you use the secret bugzilla technique of searching bugs that were marked as DUPLICATE then you often find bugs with subjects you are expecting which then point to an open bug). I'm not that convinced that this change will stop the dups but I'll make an exception this once and apologise for spamming people.
*** Bug 154690 has been marked as a duplicate of this bug. ***
*** Bug 155979 has been marked as a duplicate of this bug. ***
*** Bug 154672 has been marked as a duplicate of this bug. ***
Was this fixed in 2.6.11-1.27_FC3 ?
yes
*** Bug 157275 has been marked as a duplicate of this bug. ***