Description of problem: Starting cpuspeed, more often than not, on my Athlon64 X2 system, produces the following diagnostic: Call Trace: [<ffffffff8026e2a9>] show_trace+0xaa/0x23d [<ffffffff8026e451>] dump_stack+0x15/0x17 [<ffffffff802a655c>] print_circular_bug_tail+0x6c/0x77 [<ffffffff802a7b61>] __lock_acquire+0x853/0xa54 [<ffffffff802a8303>] lock_acquire+0x4b/0x69 [<ffffffff80265c39>] __mutex_lock_slowpath+0xe5/0x261 [<ffffffff80265ddf>] mutex_lock+0x2a/0x2e [<ffffffff8040e6a7>] store_scaling_governor+0x14f/0x19c [<ffffffff80274fb4>] store+0x4c/0x66 [<ffffffff80309b19>] sysfs_write_file+0xd1/0x103 [<ffffffff80217148>] vfs_write+0xcf/0x175 [<ffffffff80217a30>] sys_write+0x47/0x70 [<ffffffff8025ff4e>] system_call+0x7e/0x83 Version-Release number of selected component (if applicable): kernel-2.6.17-1.2416.fc6 cpuspeed-1.2.1-1.36.fc6.1 How reproducible: Not sure it's every time, but it's quite consistent during boot up Steps to Reproduce: 1.service cpuspeed start Additional info: selinux is disabled, in case the policy above has to do with it. After the failure, service cpuspeed status says: cpuspeed dead but subsys locked and indeed: root 2482 0.0 0.0 3744 280 ? Ss 04:22 0:00 cpuspeed -d -n root 2483 0.0 0.0 0 0 ? Z 04:22 0:00 [cpuspeed] <defunct> but frequency scaling still works, possibly because both cpus scale at the same time? I wonder if this has to do with the freezes I've experienced on x86_64 SMP lately... One of the times it happens most often is at 4 am, when lots of cron-started processes come up, presumably requiring more cpu activity than an otherwise mostly-idle box.
*** This bug has been marked as a duplicate of 197803 ***